1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
47 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
48 being searched. Previously we only had one for IPv4 addresses. Per the
49 documentation, the error results by default in a no-match result for the
50 list. It is logged if the unknown_in_list log_selector is used.
52 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
53 the routing operation. Previously it would silently stop routing the
56 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
57 info-request event close in time to a scheduled run timer could result in
58 the latter being missed, and no further queue scheduled runs being
59 initiated. This ouwld be more likely on high-load systems.
61 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
62 LF-only mode (as detected from the first header line). Previously we did
63 accept that in (normal) CRLF mode; this has been raised as a possible
64 attack scenario (under the name "smtp smuggling").
66 JH/13 Add an fdatasync call for the received message data file in spool, before
67 loggging reception and sending the SMTP ack. Previously we only flushed
68 the stdio buffer so there was still the possibility of a disk error.
70 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
71 address. Previously, for the last address in a header line (commonly
72 there is only one) the terminating newline was part of the logged
75 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
76 header, when parsing for address-qualification. Previously one was only
77 written if there were rewrite rules.
79 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
82 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
83 the multiple-queue-runners facility for 4.97 broke this, giving only a
84 one-time run of the queue.
86 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
89 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
90 TLS session resulted in a forced-fail, a misleading error was logged for
91 the expansino of tls_certificates. This would affect the common case of
92 that option being set (main-section options) but not having any variable
93 parts. It could also potentially affect tls_privatekeys. The underlyding
94 coding errors go back to 4.90 but were only exposed in 4.97.
96 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
97 pcre2 library starting allocating 20kB rather than 112 bytes per match
98 call, which broke the 2GB total limitation on Exim's memory management
99 when a user had over 104207 messages stored and the appendfile
100 maildir_quota_directory_regex option is in use. Release the allocated
101 memory every thosand files to avoid this.
102 The same issue arises with the ACL regex condition, which is applied
103 to every line of a received message.
105 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
106 which all recipients had been handled (itself an issue) a null-pointer
107 deref was done on trying to write a retry record. Fix that by counting
108 the outstanding recipients before trying to transmit the message.
109 The situation arose for a second MX try within a transport run, when the
110 first had perm-rejected a recipient (the only one for the connection, in
111 the case seen) during pipelining, and then closed the TCP connection.
112 The transport classified that as an I/O error, leaving the message
113 outstanding but having marked up the recipient as dealt-with. It then
114 tried another MX because of the I/O error. Fix this by converting the
115 message-level status to ok if there was a close but all recipients were
116 dealt with. Thanks to Wolfgand Breyha for debug runs.
118 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
119 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
121 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
123 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
124 option specified a fixed address, later rewriting on that address would
125 trip on the configuration data being readonly. Instead of modifying
126 in-place, copy data. Found and fixed by Peter Benie.
128 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
129 data a null pointer was followed. Find and testcase by Sebastian Bugge.
131 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
132 decimal 127 chars were not encoded, and lowercase hex was used for
133 encoded values. Outstanding since at least 1999.
139 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
140 SMTP connection" log lines.
142 JH/02 Option default value updates:
143 - queue_fast_ramp (main) true (was false)
144 - remote_max_parallel (main) 4 (was 2)
146 JH/03 Cache static regex pattern compilations, for use by ACLs.
148 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
149 Make the rewrite never match and keep the logging. Trust the
150 admin to be using verify=header-syntax (to actually reject the message).
152 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
153 (under Linux) we watch the dir containing the final file; previously
154 it would be the dir with the first symlink. We still do not monitor
157 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
158 dn_expand() is happy to pass them through.
160 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
161 the file with the proof had an unchanged name, the new proof(s) were
162 loaded on top of the old ones (and nover used; the old ones were stapled).
164 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
165 more than one message arrived in a single connection a reference from
166 the earlier message could be re-used. Often a sigsegv resulted.
167 These variables were introduced in Exim 4.87.
168 Debug help from Graeme Fowler.
170 JH/09 Fix ${filter } for conditions that modify $value. Previously the
171 modified version would be used in construction the result, and a memory
174 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
175 Find and fix by Jasen Betts.
177 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
178 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
179 the systemwide configuration to override the Exim config.
181 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
182 API changes in libopendmarc.
184 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
185 pid 1, in the normal "background daemon" mode, having to drop process-
186 group leadership also lost track of needing to create listener sockets.
188 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
189 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
191 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
192 a capture group which obtained no text (eg. "(abc)*" matching zero
193 occurrences) could cause a segfault if the corresponding $<n> was
196 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
197 included a close-brace character (eg. it itself used an expansion) an
200 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
201 starting TLS. Previously it was after, meaning that attackers on such
202 ports had to be screened using the host_reject_connection main config
203 option. The new sequence aligns better with the STARTTLS behaviour, and
204 permits defences against crypto-processing load attacks, even though it
205 is strictly an incompatible change.
206 Also, avoid sending any SMTP fail response for either the connect ACL
207 or host_reject_connection, for TLS-on-connect ports.
209 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
210 Previously this was not permitted, but it makes reasonable sense.
211 While there, restore a restriction on using it from a connect ACL; given
212 the change JH/16 it could only return false (and before 4.91 was not
215 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
216 was exactly sized compared to the log buffer, a crash occurred with the
217 misleading message "bad memory reference; pool not found".
218 Found and traced by Jasen Betts.
220 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
221 dns_again_means_nonexist included an element causing a DNS lookup which
222 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
223 included (though probably not limited to) a process crash from stack
224 memory limit, or from excessive open files. Replace this with a paniclog
225 whine (as this is likely a configuration error), and returning
228 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
229 this always failed, probably leading to the usual downgrade to in-clear
232 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
233 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
234 to not checking that list for these lookups.
236 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
239 JH/23 Fix crash in string expansions. Previously, if an empty variable was
240 immediately followed by an expansion operator, a null-indirection read
241 was done, killing the process.
243 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
244 include an SMTP response string which is longer than that supported
245 by the delivering transport. Alleviate by wrapping such lines before
248 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
249 chars (RFC limit). Previously a limit of 12 items was made, which with
250 a not-impossible References: in the message being bounced could still
251 be over-large and get stopped in the transport.
253 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
254 close. Previously a bare socket close was done.
256 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
259 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
260 message being considered for delivery was already being handled by
261 another process, and having an SMTP connection already open, the function
262 to close it tried to use an uninitialized variable. This would afftect
263 high-volume sites more, especially when running mailing-list-style loads.
264 Pollution of logs was the major effect, as the other process delivered
265 the message. Found and partly investigated by Graeme Fowler.
267 JH/29 Change format of the internal ID used for message identification. The old
268 version only supported 31 bits for a PID element; the new 64 (on systems
269 which can use Base-62 encoding, which is all currently supported ones
270 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
271 and must use Base-36). The new ID is 23 characters rather than 16, and is
272 visible in various places - notably logs, message headers, and spool file
273 names. Various of the ancillary utilities also have to know the format.
274 As well as the expanded PID portion, the sub-second part of the time
275 recorded in the ID is expanded to support finer precision. Theoretically
276 this permits a receive rate from a single comms channel of better than the
278 The major timestamp part of the ID is not changed; at 6 characters it is
279 usable until about year 3700.
280 Updating from previously releases is fully supported: old-format spool
281 files are still usable, and the utilities support both formats. New
282 message will use the new format. The one hints-DB file type which uses
283 message-IDs (the transport wait- DB) will be discarded if an old-format ID
284 is seen; new ones will be built with only new-format IDs.
285 Optionally, a utility can be used to convert spool files from old to new,
286 but this is only an efficiency measure not a requirement for operation
287 Downgrading from new to old requires running a provided utility, having
288 first stopped all operations. This will convert any spool files from new
289 back to old (losing time-precision and PID information) and remove any
290 wait- hints databases.
292 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
293 we treated them as item separators when parsing for a list item, but they
294 need to be protected by the doublequotes. While there, add handling for
297 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
298 Found and fixed by Jasen Betts. No testcase for this as my usual text
299 editor insists on emitting only valid UTF-8.
301 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
302 a null-indirection SIGSEGV for the receive process.
304 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
305 -bh use. Internal checking would spot this and take a panic.
307 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
308 In 4.96 this would expand to empty.
310 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
311 certificate. Find and fix by Andreas Metzler.
313 JH/36 Add ARC info to DMARC hostory records.
315 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
316 or fakedefer. Previously the sender could discover that the message
317 had in fact been accepted.
319 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
320 sequences. Previously the input was not noted as being tainted; notably
321 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
322 bad coding of authenticators.
324 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
325 and ${tr...}. Found and diagnosed by Heiko Schlichting.
327 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
328 could be triggered by externally-supplied input. Found by Trend Micro.
331 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
332 be triggered by externally-controlled input. Found by Trend Micro.
335 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
336 be triggered by externally-controlled input. Found by Trend Micro.
339 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
340 Make the rewrite never match and keep the logging. Trust the
341 admin to be using verify=header-syntax (to actually reject the message).
343 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
346 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
352 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
353 after reception to before a subsequent reception. This should
354 mean slightly faster delivery, and also confirmation of reception
357 JH/02 Move from using the pcre library to pcre2. The former is no longer
358 being developed or supported (by the original developer).
360 JH/03 Constification work in the filters module required a major version
361 bump for the local-scan API. Specifically, the "headers_charset"
362 global which is visible via the API is now const and may therefore
363 not be modified by local-scan code.
365 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
366 sendfile() didi not account for the way the ClamAV driver code called it.
368 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
369 time check was being done for every character; replace that with one
372 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
373 sent was prefixed with a length byte.
375 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
376 RFC 5321. Previously Dovecot (at least) would log errors during
379 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
380 in 4.61 without a reason logged. Binaries will be bigger, which might
381 matter on diskspace-constrained systems, but debug is easier.
383 JH/09 Fix macro-definition during "-be" expansion testing. The move to
384 write-protected store for macros had not accounted for these runtime
385 additions; fix by removing this protection for "-be" mode.
387 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
388 handing out large-numbered file descriptors, violating the usual Unix
389 assumption (and required by Posix) that the lowest possible number will be
390 allocated by the kernel when a new one is needed. In the daemon, and any
391 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
392 are not useable for FD_SET() [and hence select()] and overwrite the stack.
393 Assorted crashes happen.
395 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
396 main-section options or in a connect ACL, the value from the first ever
397 connection was never replaced for subsequent connections. Found by
400 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
401 though only once PCRE2 was introduced: the memory accounting used under
402 debug offset allocations by an int, giving a hard trap in early startup.
403 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
405 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
406 with underbars is given. The write-protection of configuration introduced
407 in 4.95 trapped when normalisation was applied to an option not needing
410 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
412 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
413 failing to close the certificates directory, every hour or any time it
416 JH/16 Debugging initiated by an ACL control now continues through into routing
417 and transport processes. Previously debugging stopped any time Exim
418 re-execs, or for processing a queued message.
420 JH/17 The "expand" debug selector now gives more detail, specifically on the
421 result of expansion operators and items.
423 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
424 bad comparison between the option value and the name of the file to
425 be included was done, and a mismatch was wrongly identified.
426 4.88 to 4.95 are affected.
428 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
430 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
431 supplied as the db file-pair basename. Previously, if a directory
432 path was given, for example via the autoreply "once" option, the DB
433 file.pag and file.dir files would be created in that directory's
436 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
437 "taint" log_selector. These were previously deprecated.
439 JH/22 Fix static address-list lookups to properly return the matched item.
440 Previously only the domain part was returned.
442 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
443 the call into OpenSSL to send a TLS Close was being repeated; this
444 resulted in the library waiting for the peer's Close. If that was never
445 sent we waited forever. Fix by tracking send calls.
447 JH/24 The ${run} expansion item now expands its command string elements after
448 splitting. Previously it was before; the new ordering makes handling
449 zero-length arguments simpler. The old ordering can be obtained by
450 appending a new option "preexpand", after a comma, to the "run".
452 JH/25 Taint-check exec arguments for transport-initiated external processes.
453 Previously, tainted values could be used. This affects "pipe", "lmtp" and
454 "queryprogram" transport, transport-filter, and ETRN commands.
455 The ${run} expansion is also affected: in "preexpand" mode no part of
456 the command line may be tainted, in default mode the executable name
459 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
460 the the facility was not passed across execs, and only the first message
461 passed over a connection could use BDAT; any further ones using DATA.
463 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
464 uses $sending_ip_address and an interface is specified.
465 Previously any use of the local address in the EHLO name disabled
466 PIPECONNECT, the common case being to use the rDNS of it.
468 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
469 resumption. Previously verify failed because no certificate status is
470 passed on the wire for the restarted session. Fix by using the recorded
471 ocsp status of the stored session for the new connection.
473 JH/29 TLS resumption: the key for session lookup in the client now includes
474 more info that a server could potentially use in configuring a TLS
475 session, avoiding oferring mismatching sessions to such a server.
476 Previously only the server IP was used.
478 JH/30 Fix string_copyn() for limit greater than actual string length.
479 Previously the copied amount was the limit, which could result in a
480 overlapping memcpy for newly allocated destination soon after a
481 source string shorter than the limit. Found/investigated by KM.
483 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
484 close; it may be needed for a subsequent connection. This caused a
485 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
487 JH/32 Fix CHUNKING for a second message on a connection when the first was
488 rejected. Previously we did not reset the chunking-offered state, and
489 erroneously rejected the BDAT command. Investigation help from
492 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
493 an empty address. Previously the expansion returned an error.
495 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
496 proxy. Previously these were misparsed, leading to paniclog entries.
502 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
503 related applications. Previously an "H" was used where available info
504 says that "M" should be, so change to match.
506 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
507 as arguments, so an implementation trying to copy these into a local
508 buffer was taking a taint-enforcement trap. Fix by using dynamically
509 created buffers. Similar fix for radius expansion condition.
511 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
512 reasonable, eg. to count headers. Fix by using dynamically created
513 buffers rather than a local. Do similar fixes for ACL actions "dcc",
514 "log_reject_target", "malware" and "spam"; the arguments are expanded
515 so could be handling tainted values.
517 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
518 broken the (no-op) support for this sendmail command. Restore it
519 to doing nothing, silently, and returning good status.
521 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
522 record path was given (or the default used) without a leading directory
523 path, an error occurred on trying to open it. Use the transport's working
526 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
527 transport. Previously it was the name on the DNS A-record; use instead
528 the head of the CNAME chain leading there (if there is one). This seems
529 to align better with RFC 6125.
531 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
532 smtp_accept_max_per_host allocated resources which were not released
533 when the limit was exceeded. This eventually crashed the daemon. Fix
534 by adding a release action in that path.
536 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
537 expanded; previously using tainted values was rejected. Fix by using
538 dynamically-created buffers.
540 JH/09 Relax restrictions on ACL verify condition needing access to message
541 headers. Previously they were only permitted in data and non-smtp ACLs;
542 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
543 not_blind, header_sender and header_names_ascii verification.
545 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
546 Previously a macro used one argument twice; when called with the
547 argument as an expression having side-effects, incorrect operation
548 resulted. Use an inlineable function.
550 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
551 held open for a verify callout. Previously this wan not accounted for
552 and a corrupt onward SMTP conversation resulted.
554 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
555 Previously they were embedded naively in the constructed address; when
556 needed, strip the quoting and quote the entire local_part.
557 Also make the inbound_srs expansion condition handle quoting.
559 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
560 excluded, not matching the documentation.
562 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
563 was given for the sqlite_dbfile a trap resulted.
565 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
566 "name" argument resulted in a trap. There is no reason to disallow such;
567 this was a coding error.
569 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
570 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
571 spent suspended, ignoring the POSIX definition. Previously we assumed
572 it did and a constant offset from real time could be used as a correction.
573 Change to using the same clock source for the start-of-message and the
574 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
575 exists, just to get a clock slightly more aligned to reality.
577 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
578 RFC says it is optional some validators care. The missing char was not
579 intended but triggered by a line-wrap alignment. Discovery and fix by
580 Guillaume Outters, hacked on by JH.
582 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
583 name being quoted was tainted a trap would be taken. Fix by using
584 dynamically created buffers. The routine could have been called by a
585 rewrite with the "h" flag, by using the "-F" command-line option, or
586 by using a "name=" option on a control=submission ACL modifier.
588 JH/19 SPF: change the Authentication-Results expansion component to give
589 smtp.helo when the sender domain is empty. Previously it gave
592 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
593 not in 127.0.0.0/8 to help in spotting list domains taken over by a
594 domain-parking registrar.
596 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
597 Previously when a whitespace character was specified it was not inserted
598 after removing the newline.
600 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
601 the domain part of the recipient address. This overrides any tls_sni
602 option set, which was previously used.
604 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
607 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
608 is_tainted() had an off-by-one error in the overenthusiastic direction.
609 Find and fix by Gavan. Although NetBSD is not a supported platform for
610 4.94 this bug could affect other platforms.
612 PP/01 Fix default prime selection to be consistent.
613 One path used ike23 still, instead of exim.dev.20160529.3; now both
614 execution flows will use the same DH primes (currently
615 exim.dev.20160529.3).
617 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
618 option in smtp transport, to match the documentation. Previously
619 verification was not being done in some cases where it should have been.
621 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
622 than one server was defined and depending on the platform memory layout
623 details, an internal consistency trap could be hit while walking the list
626 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
627 files. The value is used by the authresults expansion item. Previously
628 if this was used in a router or transport, a crash could result.
630 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
631 have a sufficient incidence of EINTR returns from select that an
632 interaction with other operations done by the main daemon loop exposed
633 a bug in the error-handling. This was benign apart from the log
636 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
637 for consistency with delivered "=>" and failed "**" lines. While we're
638 there, handle PRX and TFO.
640 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
641 applied. This resulted, if any header-line rewrite rules were configured,
642 in a panic-log triggerable by sending a message with a long address in
643 a header. Fix by increasing the arbitrary limit to larger than a single
644 (dewrapped) 5322 header line maximum size.
646 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
647 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
648 RFC 6648 which deprecates X- options in protocols as a general practice.
649 Changeover between the implementations is handled by the mechanisms
652 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
653 a remote which uses callout/hold. Previously the local was queued.
655 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
658 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
659 log line. The asterisk is now consistently appended to the remote IP
660 (and port, if given), and will also be provided on defer and fail log
661 lines. Previously it could be placed on the local IP if that was being
662 logged, and was only provided on delivery lines.
664 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
666 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
667 authenticator client_send option. Previously the next char, after a pair
668 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
669 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
670 documentation. There is still no way to get a leading ^ immediately
671 after a NUL (ie. for the password of a PLAIN method authenticator.
673 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
674 files. For bad sizes read, delete the record and whine to paniclog.
676 JH/38 When logging an AUTH failure, as server, do not include sensitive
677 information. Previously, the credentials would be included if given
678 as part of the AUTH command line and an ACL denied authentication.
680 JH/39 Bug 2691: fix $local_part_data. When the matching list element
681 referred to a file, bad data was returned. This likely also affected
684 JH/40 The gsasl authenticator now supports caching of the salted password
685 generated by the client-side implementation. This required the addition
686 of a new variable: $auth4.
688 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
689 left undeleted; the attempt to re-create it then failed - resulting in
690 the usual "SIGHUP tp have daemon reload configuration" to not work.
691 This affected any platform not supporting "abstract" Unix-domain
692 sockets (i.e. not Linux).
694 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
695 recipients" response to RCPT in a later response, with a 250. The
696 previous coding assumed this would not happen, and under PIPELINING
697 would result in both lost and duplicate recipients for a message.
699 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
700 Previously the weighting was incorrectly applied. Similar fix for socks
701 proxies. Found and fixed by Heiko Schlichting.
703 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
704 not handle sub-lists included using the +namedlist syntax. While
705 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
706 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
709 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
710 Previously the permanent pool was used, so the store could not be freed.
711 This meant a connection with many messages would use continually-growing
714 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
715 per-pool so as not to waste too much space. Previously a constant size
716 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
717 DOS attacks harder. The cost is wasted memory use in the larger blocks.
719 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
720 same space repeatedly, and a smaller process footprint.
722 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
723 for GnuTLS. Previously the name started "%s" which, while not a bug,
724 looked as if if might be one.
726 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
727 "continued connection") make the $proxy_* variables available. Previously
728 the information was not passed across the exec() call for subsequent
729 transport executions. This also mean that the log lines for the
730 messages can show the proxy information.
732 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
733 receive time. With modern systems the difference is significant.
734 The historical behaviour can be restored by disabling (a new) log_selector
735 "queue_time_exclusive".
737 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
738 content) were specifically tested for. Now, also cover expansions
739 resulting in ACL names and inline ACL content.
741 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
742 making it unusable in complex expressions.
744 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
745 fail with a taint-check on the spoolfile name, and leave the message
748 HS/01 Enforce absolute PID file path name.
750 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
752 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
754 PP/02 Bug 2643: Correct TLS DH constants.
755 A missing NUL termination in our code-generation tool had led to some
756 incorrect Diffie-Hellman constants in the Exim source.
757 Reported by kylon94, code-gen tool fix by Simon Arlott.
759 PP/03 Impose security length checks on various command-line options.
760 Fixes CVE-2020-SPRSS reported by Qualys.
762 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
763 better. Reported by Qualys.
765 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
766 providing a particularly obnoxious sender full name.
769 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
771 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
774 PP/08 Change default for recipients_max from unlimited to 50,000.
776 PP/09 Fix security issue with too many recipients on a message (to remove a
777 known security problem if someone does set recipients_max to unlimited,
778 or if local additions add to the recipient list).
779 Fixes CVE-2020-RCPTL reported by Qualys.
781 PP/10 Fix security issue in SMTP verb option parsing
782 Fixes CVE-2020-EXOPT reported by Qualys.
784 PP/11 Fix security issue in BDAT state confusion.
785 Ensure we reset known-good where we know we need to not be reading BDAT
786 data, as a general case fix, and move the places where we switch to BDAT
787 mode until after various protocol state checks.
788 Fixes CVE-2020-BDATA reported by Qualys.
790 HS/03 Die on "/../" in msglog file names
792 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
793 the Exim runtime user are allowed to create files.
795 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
798 QS/03 When reading the output from interpreted forward files we do not
799 pass the pipe between the parent and the interpreting process to
800 executed child processes (if any).
802 QS/04 Always die if requested from internal logging, even is logging is
805 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
806 the API; compilation noo longer completes with DMARC support included.
807 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
808 been reported on other platforms.
810 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
812 JH/56 Make the majority of info read from config files readonly, for defence-in-
813 depth against exploits. Suggestion by Qualys.
814 Not supported on Solaris 10.
816 JH/57 Fix control=fakreject for a custom message containing tainted data.
817 Previously this resulted in a log complaint, due to a re-expansion present
818 since fakereject was originally introduced.
820 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
821 modifier was given, a loop resulted.
823 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
824 pipelined SMTP command followed the BDAT LAST then it would be
825 incorrectly treated as part of the message body, causing a verification
828 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
829 require looser validation rules than those for 821-level addresses,
830 which only permit IP addresses.
836 JH/01 Avoid costly startup code when not strictly needed. This reduces time
837 for some exim process initialisations. It does mean that the logging
838 of TLS configuration problems is only done for the daemon startup.
840 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
842 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
843 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
846 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
847 DKIM signing is being done. Previously a transport_filter would always
848 disable CHUNKING, falling back to traditional DATA.
850 JH/05 Regard command-line recipients as tainted.
852 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
854 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
855 PAM library frees one of the arguments given to it, despite the
856 documentation. Therefore a plain malloc must be used.
858 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
859 on-stack buffers were used, resulting in a taint trap when DSN information
860 copied from a received message was written into the buffer.
862 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
863 the ordering of its ARC headers. This caused a crash.
865 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
866 a new record was being constructed with information from the peer, a trap
869 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
870 installation would get error messages from DMARC verify, when it hit the
871 nonexistent file indicated by the default. Distros wanting DMARC enabled
872 should both provide the file and set the option.
873 Also enforce no DMARC verification for command-line sourced messages.
875 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
876 could, depending on the platform, hang at the STARTTLS response.
878 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
879 message on a connection. Previously if one message had ARC headers and
880 the following one did not, a crash could result when adding an
881 Authentication-Results: header.
883 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
884 Exim main code and Exim-related utities. The introduction of taint
885 tracking also did many adjustments to string handling. Since then, eximon
886 frequently terminated with an assert failure.
888 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
889 check for 452 responses. This slightly helps the inefficieny of doing
890 a large alias-expansion into a recipient-limited target. The max_rcpt
891 transport option still applies (and at the current default, will override
892 the new feature). The check is done for either cause of synch, and forces
893 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
894 connection. The new facility is not tunable at this time.
896 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
897 library live data was being used, so the results became garbage. Make
898 copies while it is still usable.
900 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
901 on delivery deferred (==) and failed (**) lines (if a delivery was
902 attemtped). Previously it was only on completion (=>) lines.
904 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
905 for the expansion of the server_scram_iter and server_scram_salt options.
907 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
908 are now specifically given a NO_DATA response without hitting the system
909 resolver. The library goes on to do the now-standard TXT lookup.
910 Use of dnsdb lookups is not affected.
912 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
913 only retrieve the errormessage once. Previously two calls to dlerror()
914 were used, and the second one (for mainlog/paniclog) retrieved null
917 JH/20 Taint checking: disallow use of tainted data for
918 - the appendfile transport file and directory options
919 - the pipe transport command
920 - the autoreply transport file, log and once options
921 - file names used by the redirect router (including filter files)
923 - paths used by single-key lookups
924 Previously this was permitted.
926 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
927 adjusted the size of a major service buffer; this failed because the
928 buffer was in use at the time. Change to a compile-time increase in the
929 buffer size, when this authenticator is compiled into exim.
931 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
932 previous fast-mode was untenable in the face of glibs using mmap to
933 support larger malloc requests.
935 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
936 New values supported, if defined on system where compiled:
937 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
938 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
940 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
941 running a limited number of proceses in parallel, a benefit is gained. The
942 amount varies with the platform hardware and load. The use of the option
943 queue_run_in_order means we cannot do this, as ordering becomes
946 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
947 had introduced a string-copy (for ensuring NUL-termination) which was not
948 appropriate for that case, which can include embedded NUL bytes in the
949 block of data. Investigation showed the copy to actually be needless, the
950 data being length-specified.
952 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
953 done during a receiving connection, and both used TLS, global info was
954 used rather than per-connection info for tracking the state of data
955 queued for transmission. This could result in a connection hang.
957 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
958 Previously, when delivering serveral messages down a single connection
959 only the first would provide a SIZE. This was due to the size information
960 not being properly tracked.
962 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
963 TAI (at 37 seconds currently), pretend to be in UTC for time-related
964 expansion and logging. Previously, spurious values such as a future
965 minute could be seen.
967 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
968 it could crash from a null-deref. This could also affect the
969 ${addresses: } operator and ${readsock } item.
971 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
972 message following a mime one, the variable was not reset.
974 JH/30 When an pipelined-connect fails at the first response, assume incorrect
975 cached capability (perhaps the peer reneged?) and immediately retry in
978 JH/31 Fix spurious detection of timeout while writing to transport filter.
980 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
981 an attempt to copy the string was made before checking it.
983 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
984 taint of the lookup key was maintained; we now regard the presence in the
985 filesystem as sufficient validation.
987 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
988 argument is supplied.
990 JH/35 The dsearch lookup now requires that the directory is an absolute path.
991 Previously this was not checked, and nonempty relative paths made an
992 access under Exim's current working directory.
994 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
995 Previously no event was raised.
997 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
998 parameter supplied by the sender MAIL FROM command. Previously it was
999 ignored, and only the check_spool_space option value for the required
1002 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1003 the size of the signing public-key. Previously it was instead giving
1004 the size of the signature hash.
1006 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1007 the default. See the (new) dkim_verify_min_keysizes option.
1009 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1010 and an ACL use a lookup for checking either the local_part or domain,
1011 stale data could be accessed. Ensure that variable references are
1012 dropped between messages.
1014 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1015 by the client was not checked as pointing within response data before
1016 being used. A malicious client could thus cause an out-of-bounds read and
1017 possibly gain authentication. Fix by adding the check.
1019 JH/42 Internationalisation: change the default for downconversion in the smtp
1020 transport to be "if needed". Previously it was "as previously set" for
1021 the message, which usually meant "if needed" for message-submission but
1022 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1023 even when the envelope addresses did not need it, resulting in forwarding
1024 failures to non-supporting MTAs. A downconvert in such cases will be
1025 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1026 transport. The change does mean that addresses needing conversion will
1027 be converted when previously a delivery failure would occur.
1029 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1030 response was received it would be used unchecked in a fail-DSN, violating
1031 standards on line-length limits. Truncate if needed.
1033 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1034 doesn't work. (Additionally add a new main config option to configure the
1041 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1042 side, to decode a TLS 1.3 packet capture.
1044 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1045 Previously the default library behaviour applied, sending two, each in
1046 its own TCP segment.
1048 JH/03 Debug output for ACL now gives the config file name and line number for
1051 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1053 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1055 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1056 buffer overrun for (non-chunking) other transports.
1058 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1059 TLS1.3, means that a server rejecting a client certificate is not visible
1060 to the client until the first read of encrypted data (typically the
1061 response to EHLO). Add detection for that case and treat it as a failed
1062 TLS connection attempt, so that the normal retry-in-clear can work (if
1063 suitably configured).
1065 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1066 and/or domain. Found and fixed by Jason Betts.
1068 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1069 configuration). If a CNAME target was not a wellformed name pattern, a
1072 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1073 the OS reports them interleaved with other addresses.
1075 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1076 used both for input and for a verify callout, both encrypted, SMTP
1077 responses being sent by the server could be lost. This resulted in
1078 dropped connections and sometimes bounces generated by a peer sending
1081 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1082 string. Previously it was possible to cause undefined behaviour in a
1083 library routine (usually a crash). Found by "zerons".
1085 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1088 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1089 API was removed, so update to use the newer ones.
1091 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1092 any timeout set, is taking a long time. Previously we would hang on to a
1093 rotated logfile "forever" if the input was arriving with long gaps
1094 (a previous attempt to fix addressed lack, for a long time, of initial
1097 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1098 shared (NFS) environment. The length of the tempfile name is now
1099 4 + 16 ("hdr.$message_exim_id") which might break on file
1100 systems which restrict the file name length to lower values.
1101 (It was "hdr.$pid".)
1103 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1104 shared (NFS) environment.
1106 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1107 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1110 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1111 on some platforms for bit 31.
1113 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1114 to changes apparently associated with TLS1.3 handling some of the APIs
1115 previously used were either nonfunctional or inappropriate. Strings
1116 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1117 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1118 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1119 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1120 and the use of specific cipher names in the encrypted= ACL condition.
1122 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1124 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1125 verification result was not updated unless hosts_require_ocsp applied.
1127 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1128 queue_list_requires_admin set to false, non-admin users were denied the
1131 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1132 directory-of-certs mode. Previously they were advertised despite the
1135 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1136 A single TCP connection by a client will now hold a TLS connection open
1137 for multiple message deliveries, by default. Previously the default was to
1140 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1141 default. If built with the facility, DANE will be used. The facility
1142 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1144 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1145 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1146 must be defined and you must still, unless you define DISABLE_TLS, manage
1147 the the include-dir and library-file requirements that go with that
1148 choice. Non-TLS builds are still supported.
1150 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1153 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1154 default. If the platform supports and has the facility enabled, it will
1155 be requested on all coneections.
1157 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1158 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1160 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1162 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1163 success-DSN messages. Previously the From: header was always the default
1164 one for these; the option was ignored.
1166 JH/28 Fix the timeout on smtp response to apply to the whole response.
1167 Previously it was reset for every read, so a teergrubing peer sending
1168 single bytes within the time limit could extend the connection for a
1169 long time. Credit to Qualsys Security Advisory Team for the discovery.
1171 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1172 delivery address, which leaked information of the results of local
1173 forwarding. Change to the original envelope recipient address, per
1176 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1177 requested. Previously not bounce was generated and a log entry of
1178 error ignored was made.
1180 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1182 JH/32 Introduce a general tainting mechanism for values read from the input
1183 channel, and values derived from them. Refuse to expand any tainted
1184 values, to catch one form of exploit.
1186 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1187 was unused and the unexpanded text used for the test. Found and
1188 fixed by Ruben Jenster.
1190 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1191 an attempt to use a TLS library read routine dereffed a nul pointer,
1194 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1195 them in our smtp response.
1197 JH/36 Have the general router option retry_use_local_part default to true when
1198 any of the restrictive preconditions are set (to anything). Previously it
1199 was only for check_local user. The change removes one item of manual
1200 configuration which is required for proper retries when a remote router
1201 handles a subset of addresses for a domain.
1203 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1204 link count into consideration.
1206 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1207 caused the extension of big_buffer, the following lines were ignored.
1209 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1210 accordance with RFC 2308. Previously there was no expiry, so a longlived
1211 receive process (eg. due to ACL delays) versus a short SOA value could
1214 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1216 JH/39 Promote DMARC support to mainline.
1218 JH/40 Bug 2452: Add a References: header to DSNs.
1220 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1221 parameters. The relevant library call is documented as "Deprecated: This
1222 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1223 3.6.0, DH parameters are negotiated following RFC7919."
1225 HS/06 Change the default of dnssec_request_domains to "*"
1227 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1228 carried on and emitted a BDAT command, even when PIPELINING was not
1231 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1232 buffer was used for the filename, resulting in a trap when tainted
1233 arguments (eg. $domain) were used.
1235 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1236 recommended to avoid a possible server-load attack. The feature can be
1237 re-enabled via the openssl_options main cofiguration option.
1239 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1240 for version 4.90 - adding a "more data" boolean to the arguments.
1241 Bumped the ABI version number also, this having been missed previously;
1242 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1243 respect of usage of smtp_printf() by either local_scan code or libraries
1244 accessed via the ${dlfunc } expansion item. Both will need coding
1245 adjustment for any calls to smtp_printf() to match the new function
1246 signature; a FALSE value for the new argument is always safe.
1248 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1249 the file-offset (which the Linux syscall does, and exim expects); this
1250 resulted in an indefinite loop.
1252 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1253 to do ARC verification. The Authentication-Results: header line added
1254 by the configuration then had no ARC item.
1256 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1257 an incorrect mode was used for reading the result, resulting in it being
1264 JH/01 Remove code calling the customisable local_scan function, unless a new
1265 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1267 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1268 non-signal-safe functions being used.
1270 JH/03 Bug 2269: When presented with a received message having a stupidly large
1271 number of DKIM-Signature headers, disable DKIM verification to avoid
1272 a resource-consumption attack. The limit is set at twenty.
1274 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1275 report of oldest_pass in ${authres } in consequence, and separate out
1276 some descriptions of reasons for verification fail.
1278 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1279 files in the spool were present and unlocked. A queue-runner could spot
1280 them, resulting in a duplicate delivery. Fix that by doing the unlock
1281 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1282 add more error-checking on spoolfile handling while that code is being
1285 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1286 No known attacks, no CVE, this is defensive hardening.
1288 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1289 a queue-runner could start a delivery while other operations were ongoing.
1290 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1291 Found and investigated by Tim Stewart. Fix by using the open message data
1292 file handle rather than opening another, and not locally closing it (which
1293 releases a lock) for that case, while creating the temporary .eml format
1294 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1296 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1297 $sender_verify_failure/$recipient_verify_failure to "random".
1299 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1302 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1303 Previously this would segfault.
1305 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1308 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1309 like zero, since the resolver should be doing this for us, But we need one
1310 as a CNAME but no MX presence gets the CNAME returned; we need to check
1311 that doesn't point to an MX to declare it "no MX returned" rather than
1312 "error, loop". A new main option is added so the older capability of
1313 following some limited number of chain links is maintained.
1315 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1317 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1318 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1319 now-deprecated earlier definitions used only specified the range up to TLS
1320 1.2 (in the older-version library docs).
1322 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1324 JH/15 Rework TLS client-side context management. Stop using a global, and
1325 explicitly pass a context around. This enables future use of TLS for
1326 connections to service-daemons (eg. malware scanning) while a client smtp
1327 connection is using TLS; with cutthrough connections this is quite likely.
1329 JH/16 Fix ARC verification to do AS checks in reverse order.
1331 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1333 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1334 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1335 Previously the "utf8" would be re-prepended for every additional message.
1337 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1338 Previously thery were accepted, resulting in issues when attempting to
1339 forward messages to a non-supporting MTA.
1341 PP/02 Let -n work with printing macros too, not just options.
1343 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1344 one parent address was copied, and bogus data was used at delivery-logging
1345 time. Either a crash (after delivery) or bogus log data could result.
1346 Discovery and analysis by Tim Stewart.
1348 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1349 Previously if the string ended mid-character, we did not insert the
1350 promised '?' replacement.
1352 PP/04 Documentation: current string operators work on bytes, not codepoints.
1354 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1355 should pack well giving a smaller memory footprint so better caching and
1356 therefore performance. Group the declarations where this can't be done so
1357 that the byte-sized flag variables are not interspersed among pointer
1358 variables, giving a better chance of good packing by the compiler.
1360 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1361 non-null, to avoid issues with sites running BATV. Previously reports were
1362 sent with an empty envelope sender so looked like bounces.
1364 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1365 The ignore_error flag wasn't being returned from the filter subprocess so
1366 was not set for later routers. Investigation and fix by Matthias Kurz.
1368 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1369 and a msg:complete for the whole, when a message is manually removed using
1370 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1372 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1373 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1374 hope that is portable enough.
1376 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1377 requiring a known-CA anchor certificate; make it now rely entirely on the
1378 TLSA as an anchor. Checking the name on the leaf cert against the name
1379 on the A-record for the host is still done for TA (but not for EE mode).
1381 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1382 included in delivery lines for non-proxied connections, when compiled with
1383 SUPPORT_SOCKS and running with proxy logging enabled.
1385 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1386 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1387 move the existing event to fire before the normal logging of message
1388 failure so that custom logging is bracketed by normal logging.
1390 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1391 msg:fail:internal event. Developement by Matthias Kurz.
1393 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1394 far too small for todays use of crypto signatures stored there. Go all
1395 the way to the max DNS message size of 64kB, even though this might be
1396 overmuch for IOT constrained device use.
1398 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1399 copy a string over itself. The library routine is documented as not
1400 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1402 JH/32 For main options check_spool_space and check_inode_space, where the
1403 platform supports 64b integers, support more than the previous 2^31 kB
1404 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1405 the previous G, M, k.
1407 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1408 $authenticated_fail_id variable on authentication failure. Previously
1411 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1412 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1413 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1414 GNUTLS_SEC_PARAM_MEDIUM.
1416 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1417 side. Previously we would continue as if no SNI had been received.
1419 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1420 "<" character, which should be a list-separator specification, we walked
1421 off past the nul-terimation.
1423 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1424 causes) even when the retry time is not yet met. Previously they were
1425 not, meaning that when (say) an account was over-quota and temp-rejecting,
1426 and multiple senders' messages were queued, only one sender would get
1427 notified on each configured delay_warning cycle.
1429 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1431 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1432 cipher-suites, an error can be left on the stack even for a succeeding
1433 accept; this results in impossible error messages when a later operation
1436 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1437 return error codes indicating retry. Under TLS1.3 this becomes required.
1439 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1440 it only wrote the new authenticators, resulting in a lack of tracking of
1441 peer changes of ESMTP extensions until the next cache flush.
1443 JH/41 Fix the loop reading a message header line to check for integer overflow,
1444 and more-often against header_maxsize. Previously a crafted message could
1445 induce a crash of the recive process; now the message is cleanly rejected.
1447 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1448 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1454 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1455 When redis_servers is set to a list of > 1 element, and the Redis servers
1456 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1457 case of MOVED into a DEFER case instead, thus moving the query onto the
1458 next server in the list. For a cluster of N elements, all N servers must
1459 be defined in redis_servers.
1461 GF/02 Catch and remove uninitialized value warning in exiqsumm
1462 Check for existence of @ARGV before looking at $ARGV[0]
1464 JH/01 Replace the store_release() internal interface with store_newblock(),
1465 which internalises the check required to safely use the old one, plus
1466 the allocate and data copy operations duplicated in both (!) of the
1467 extant use locations.
1469 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1470 modifier. This matches the restriction on the commandline.
1472 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1473 Previously only the last row was returned.
1475 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1476 we assumed that tags in the header were well-formed, and parsed the
1477 element content after inspecting only the first char of the tag.
1478 Assumptions at that stage could crash the receive process on malformed
1481 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1482 While running the DKIM ACL we operate on the Permanent memory pool so that
1483 variables created with "set" persist to the DATA ACL. Also (at any time)
1484 DNS lookups that fail create cache records using the Permanent pool. But
1485 expansions release any allocations made on the current pool - so a dnsdb
1486 lookup expansion done in the DKIM ACL releases the memory used for the
1487 DNS negative-cache, and bad things result. Solution is to switch to the
1488 Main pool for expansions.
1489 While we're in that code, add checks on the DNS cache during store_reset,
1490 active in the testsuite.
1491 Problem spotted, and debugging aided, by Wolfgang Breyha.
1493 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1494 When none of the hosts presented to a transport match an already-open
1495 connection, close it and proceed with the list. Previously we would
1496 queue the message. Spotted by Lena with Yahoo, probably involving
1499 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1500 Previously a spurious "250 OK id=" response was appended to the proper
1503 JH/08 The "support for" informational output now, which built with Content
1504 Scanning support, has a line for the malware scanner interfaces compiled
1505 in. Interface can be individually included or not at build time.
1507 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1508 by the template makefile "src/EDITME". The "STREAM" support for an older
1509 ClamAV interface method is removed.
1511 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1512 rows affected is given instead).
1514 JH/11 The runtime Berkeley DB library version is now additionally output by
1515 "exim -d -bV". Previously only the compile-time version was shown.
1517 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1518 SMTP connection. Previously, when one had more recipients than the
1519 first, an abortive onward connection was made. Move to full support for
1520 multiple onward connections in sequence, handling cutthrough connection
1521 for all multi-message initiating connections.
1523 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1524 routers. Previously, a multi-recipient message would fail to match the
1525 onward-connection opened for the first recipient, and cause its closure.
1527 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1528 a timeout on read on a GnuTLS initiating connection, resulting in the
1529 initiating connection being dropped. This mattered most when the callout
1530 was marked defer_ok. Fix to keep the two timeout-detection methods
1533 JH/15 Relax results from ACL control request to enable cutthrough, in
1534 unsupported situations, from error to silently (except under debug)
1535 ignoring. This covers use with PRDR, frozen messages, queue-only and
1538 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1540 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1541 metadata, resulting in a crash in free().
1543 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1544 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1545 Broken also in d185889f4, with init system revamp.
1547 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1548 Previously we abruptly closed the connection after reading a malware-
1549 found indication; now we go on to read the "scan ok" response line,
1552 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1553 ACL. Previously, a crash would result.
1555 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1556 macro text after a replacement (previously it was only once per line) and
1557 by skipping builtin macros when searching for an uppercase lead character.
1559 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1560 for the build is renamed.
1562 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1563 was allocated for every new TLS startup, meaning one per message. Fix
1564 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1566 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1567 reported the original. Fix to report (as far as possible) the ACL
1568 result replacing the original.
1570 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1571 OpenSSL. Certificate information is loaded for every new TLS startup,
1572 and the resources needed to be freed.
1574 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1576 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1579 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1580 DATA response info to the (existing) per-recipient response info for
1581 the "C=" log element. It can have useful tracking info from the
1582 destination system. Patch from Simon Arlott.
1584 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1585 length value. Previously this would segfault.
1587 HS/02 Support Avast multiline protoocol, this allows passing flags to
1588 newer versions of the scanner.
1590 JH/28 Ensure that variables possibly set during message acceptance are marked
1591 dead before release of memory in the daemon loop. This stops complaints
1592 about them when the debug_store option is enabled. Discovered specifically
1593 for sender_rate_period, but applies to a whole set of variables.
1594 Do the same for the queue-runner and queue-list loops, for variables set
1595 from spool message files. Do the same for the SMTP per-message loop, for
1596 certain variables indirectly set in ACL operations.
1598 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1599 as a multi-recipient message from a mailinglist manager). The coding had
1600 an arbitrary cutoff number of characters while checking for more input;
1601 enforced by writing a NUL into the buffer. This corrupted long / fast
1602 input. The problem was exposed more widely when more pipelineing of SMTP
1603 responses was introduced, and one Exim system was feeding another.
1604 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1605 receiving system, and refused recipients seen by the sending system
1606 (propating to people being dropped from mailing lists).
1607 Discovered and pinpointed by David Carter.
1609 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1610 replaced by the ${authresults } expansion.
1612 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1614 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1615 allows proper process termination in container environments.
1617 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1618 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1620 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1621 and "err_perm", deprecated since 4.83 when the RFC-defined words
1622 "temperror" and "permerror" were introduced.
1624 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1625 transport-filters or DKIM-signing. The restriction was lost in the
1626 consolidation of verify-callout and delivery SMTP handling.
1627 Extend the restriction to also cover ARC-signing.
1629 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1630 in defer=pass mode supply a 450 to the initiator. Previously the message
1633 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1634 tls_require_ciphers is used as before.
1636 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1637 "pass_unscanned". Only tmpfails from the scanner are written to
1638 the paniclog, as they may require admin intervention (permission
1639 denied, license issues). Other scanner errors (like decompression
1640 bombs) do not cause a paniclog entry.
1642 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1643 Previously it was possible to log spurious information in receive log
1646 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1647 triggered odd behaviour from Outlook Express clients.
1649 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1652 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1653 since the IETF WG has not yet settled on that versus the original
1654 "bare" representation.
1656 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1657 Previously the millisecond value corrupted the output.
1658 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1659 corrupted the output.
1665 JH/01 Rework error string handling in TLS interface so that the caller in
1666 more cases is responsible for logging. This permits library-sourced
1667 string to be attached to addresses during delivery, and collapses
1668 pairs of long lines into single ones.
1670 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1671 during configuration. Wildcards are allowed and expanded.
1673 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1674 This permits better logging.
1676 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1677 do not close it down and have the child transport start it up again on
1678 the passed-on TCP connection. Instead, proxy the child (and any
1679 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1680 affected: the continued delivery log lines do not have any DNSSEC, TLS
1681 Certificate or OCSP information. TLS cipher information is still logged.
1683 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1684 identical IP addresses on different listening ports. Will also affect
1687 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1688 add noisy ifdef guards to special-case this sillyness.
1689 Patch from Bernd Kuhls.
1691 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1692 than 255 are no longer allowed.
1694 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1695 Disable the session-cache too, which might reduce our load. Since we
1696 currrectly use a new context for every connection, both as server and
1697 client, there is no benefit for these.
1698 GnuTLS appears to not support tickets server-side by default (we don't
1699 call gnutls_session_ticket_enable_server()) but client side is enabled
1700 by default on recent versions (3.1.3 +) unless the PFS priority string
1703 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1704 <https://reproducible-builds.org/specs/source-date-epoch/>.
1706 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1707 the check for any unsuccessful recipients did not notice the limit, and
1708 erroneously found still-pending ones.
1710 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1711 MSG_MORE. Only in-clear (not on TLS connections).
1713 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1714 transport-filter is involved we can buffer the headers in memory for
1715 creating the signature, and read the spool data file once for the
1716 signature and again for transmission.
1718 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1719 4.77 as the kernel support then wasn't solid, having issues in 64bit
1720 mode. Now, it's been long enough. Add support for FreeBSD also.
1722 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1723 case where the routing stage had gathered several addresses to send to
1724 a host before calling the transport for the first, we previously failed
1725 to close down TLS in the old transport process before passing the TCP
1726 connection to the new process. The new one sent a STARTTLS command
1727 which naturally failed, giving a failed delivery and bloating the retry
1728 database. Investigation and fix prototype from Wolfgang Breyha.
1730 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1731 false-negatives in the check that the sender had not preempted a response
1732 or prompt from Exim (running as a server), due to that code's lack of
1733 awareness of the SMTP input buffering.
1735 PP/04 Add commandline_checks_require_admin option.
1736 Exim drops privileges sanely, various checks such as -be aren't a
1737 security problem, as long as you trust local users with access to their
1738 own account. When invoked by services which pass untrusted data to
1739 Exim, this might be an issue. Set this option in main configuration
1740 AND make fixes to the calling application, such as using `--` to stop
1743 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1744 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1745 responses to those, into a single TLS record each way (this usually means
1746 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1749 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1750 affects you only if you're dancing at the edge of the param size limits.
1751 If you are, and this message makes sense to you, then: raise the
1752 configured limit or use OpenSSL 1.1. Nothing we can do for older
1755 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1756 cmdline element to get the documented default one. Previously it was
1759 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1760 are unsuable for non-message-specific reasons - eg. connection timeout,
1761 banner-time rejection.
1763 JH/16 Fix logging of delivery remote port, when specified by router, under
1766 PP/06 Repair manualroute's ability to take options in any order, even if one
1767 is the name of a transport.
1770 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1772 JH/17 Change the list-building routines interface to use the expanding-string
1773 triplet model, for better allocation and copying behaviour.
1775 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1776 Previously it was constructed the first time a possibly-matching string
1777 was met in the configuration file input during startup; now it is done
1780 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1781 compatible one, to avoid the (poorly documented) possibility of a config
1782 file in the working directory redirecting the DB files, possibly correpting
1783 some existing file. CVE-2017-10140 assigned for BDB.
1785 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1786 cache-hot. Previously, although the result was properly cached, the
1787 initial verify call returned a defer.
1789 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1790 the main verify for receipient in uncached-mode.
1792 JH/22 Retire historical build files to an "unsupported" subdir. These are
1793 defined as "ones for which we have no current evidence of testing".
1795 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1796 if present. Previously it was ignored.
1798 JH/24 Start using specified-initialisers in C structure init coding. This is
1799 a C99 feature (it's 2017, so now considered safe).
1801 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1802 if was a fixed-sized field and bitmask ops via macros; it is now more
1805 PP/07 GitHub PR 56: Apply MariaDB build fix.
1806 Patch provided by Jaroslav Škarvada.
1808 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1809 during Coverity cleanups [4.87 JH/47]
1810 Diagnosis and fix provided by Michael Fischer v. Mollard.
1812 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1813 the right size to place the terminating semicolon on its own folded
1814 line, the header hash was calculated to an incorrect value thanks to
1815 the (relaxed) space the fold became.
1817 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1818 and confused the parent.
1820 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1821 which could crash as a result. This could lead to undeliverable messages.
1823 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1826 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1827 matters on fast-turnover and PID-randomising systems, which were getting
1828 out-of-order delivery.
1830 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1831 a possibly-overlapping copy. The symptom was that "Remote host closed
1832 connection in response to HELO" was logged instead of the actual 4xx
1835 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1836 Previously only that bufferd was discarded, resulting in SYMTP command
1839 JH/32 DKIM: when a message has multiple signatures matching an identity given
1840 in dkim_verify_signers, run the dkim acl once for each. Previously only
1841 one run was done. Bug 2189.
1843 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1844 "panic the current process" to "deliberately defer". The panic log is
1845 still written with the problem list name; the mail and reject logs now
1846 get a temp-reject line for the message that was being handled, saying
1847 something like "domains check lookup or other defer". The SMTP 451
1848 message is still "Temporary local problem".
1850 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1851 A crafted sequence of BDAT commands could result in in-use memory beeing
1852 freed. CVE-2017-16943.
1854 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1855 from SMTP input. Previously it was always done; now only done for DATA
1856 and not BDAT commands. CVE-2017-16944.
1858 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1859 to the message (such as an overlong header line). Previously this was
1860 not done and we did not exit BDAT mode. Followon from the previous item
1861 though a different problem.
1867 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1868 than -2003 did; needs libidn2 in addition to libidn.
1870 JH/02 The path option on a pipe transport is now expanded before use.
1872 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1873 Patch provided by "Björn", documentation fix added too.
1875 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1876 missing a wire-to-host endian conversion.
1878 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1879 close after a BDAT command line could be taken as a following command,
1880 giving a synch failure. Fix by only checking for synch immediately
1881 before acknowledging the chunk.
1883 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1884 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1885 macro. Patches provided by Josh Soref.
1887 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1888 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1889 by IANA as a service extension. However, John Klensin suggests that we
1892 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1893 the dkim code may be unix-mode line endings rather than smtp wire-format
1894 CRLF, so prepend a CR to any bare LF.
1896 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1897 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1899 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1900 the first were themselves being wrongly included in the feed into dkim
1901 processing; with most chunk sizes in use this resulted in an incorrect
1902 body hash calculated value.
1904 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1905 DKIM signature block, for verification. Although advised against by
1906 standards it is specifically not ruled illegal.
1908 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1910 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1911 missing a body hash (the bh= tag).
1913 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1914 It seems that HAProxy sends the Proxy Protocol information in clear and
1915 only then does a TLS startup, so do the same.
1917 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1918 TCP connections (such as for Spamd) unless the daemon successfully set
1919 Fast Open mode on its listening sockets. This fixes breakage seen on
1920 too-old kernels or those not configured for Fast Open, at the cost of
1921 requiring both directions being enabled for TFO, and TFO never being used
1922 by non-daemon-related Exim processes.
1924 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1925 endings, at least on the first header line. Try to canonify any that get
1926 past that check, despite the cost.
1928 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1929 now limited to an arbitrary five deep, while parsing addresses with the
1930 strip_excess_angle_brackets option enabled.
1932 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1933 instead leave the unprompted TLS handshake in socket buffer for the
1934 TLS library to consume.
1936 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1938 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1940 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1941 region in one operation, for speed, and this leaves assigned pointers
1942 dangling. Add checks run only under the testsuite which checks all
1943 variables at a store-reset and panics on a dangling pointer; add code
1944 explicitly nulling out all the variables discovered. Fixes one known
1945 bug: a transport crash, where a dangling pointer for $sending_ip_address
1946 originally assigned in a verify callout, is re-used.
1948 PP/06 Drop '.' from @INC in various Perl scripts.
1950 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1952 PP/08 Reduce a number of compilation warnings under clang; building with
1953 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1954 should be warning-free.
1956 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1958 HS/01 Fix portability problems introduced by PP/08 for platforms where
1959 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1961 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1962 chunk. This allows us to accept broken chunked messages. We need a more
1963 general solution here.
1965 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1966 already-broken messages in the queue.
1968 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1970 JH/19 Fix reference counting bug in routing-generated-address tracking.
1976 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1977 supports it and a size is available (ie. the sending peer gave us one).
1979 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1980 years of being deprecated). The replacements are the ACLs
1981 acl_smtp_mime and acl_not_smtp_mime.
1983 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1984 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1985 or even an in-clear connection were permitted. Now, if the host lookup
1986 was dnssec and dane was requested then the host is only used if the
1987 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1988 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1989 if one fails this test.
1990 This means that a poorly-configured remote DNS will make it incommunicado;
1991 but it protects against a DNS-interception attack on it.
1993 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1994 non-noisy when a race steals the message being considered.
1996 JH/05 If main configuration option tls_certificate is unset, generate a
1997 self-signed certificate for inbound TLS connections.
1999 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2000 in rewrites and routers.
2002 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2003 and logged a warning sing 4.83; now they are a configuration file error.
2005 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2006 (lacking @domain). Apply the same qualification processing as RCPT.
2008 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2010 JH/10 Support ${sha256:} applied to a string (as well as the previous
2013 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2014 a cutthrough deliver is pending, as we always want to make a connection.
2015 This also avoids re-routing the message when later placing the cutthrough
2016 connection after a verify cache hit.
2017 Do not update it with the verify result either.
2019 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2020 when routing results in more than one destination address.
2022 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2023 signing (which inhibits the cutthrough capability). Previously only
2024 the presence of an option was tested; now an expansion evaluating as
2025 empty is permissible (obviously it should depend only on data available
2026 when the cutthrough connection is made).
2028 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2029 the relevant preceding SMTP command did not note the pipelining mode.
2031 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2032 Previously they were not counted.
2034 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2035 as one having no matching records. Previously we deferred the message
2036 that needed the lookup.
2038 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2039 distinguished as "(=".
2041 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2042 for missing MX records. Previously it only worked for missing A records.
2044 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2046 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2047 after the data-go-ahead and data-ack. Patch from Jason Betts.
2049 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2050 even for a "none" policy. Patch from Tony Meyer.
2052 JH/22 Fix continued use of a connection for further deliveries. If a port was
2053 specified by a router, it must also match for the delivery to be
2056 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2057 When one of the recipients of a message was unsuitable for the connection
2058 (has no matching addresses), we lost track of needing to mark it
2059 deferred. As a result mail would be lost.
2061 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2063 JH/25 Decoding ACL controls is now done using a binary search; the source code
2064 takes up less space and should be simpler to maintain. Merge the ACL
2065 condition decode tables also, with similar effect.
2067 JH/26 Fix problem with one_time used on a redirect router which returned the
2068 parent address unchanged. A retry would see the parent address marked as
2069 delivered, so not attempt the (identical) child. As a result mail would
2072 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2073 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2074 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2077 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2078 Main config options check_{log,spool}_{inodes,space} are now
2079 100 inodes, 10MB unless set otherwise in the configuration.
2081 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2082 Previously it only applied to the main-section connection policy
2085 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2087 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2088 by me. Added RFC7919 DH primes as an alternative.
2090 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2091 are not in the system include path.
2093 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2094 GnuTLS, when a session startup failed (eg because the client disconnected)
2095 Exim did stdio operations after fclose. This was exposed by a recent
2096 change which nulled out the file handle after the fclose.
2098 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2099 signed directly by the cert-signing cert, rather than an intermediate
2100 OCSP-signing cert. This is the model used by LetsEncrypt.
2102 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2104 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2105 an incoming connection.
2107 HS/02 Bug 1802: Do not half-close the connection after sending a request
2110 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2111 fallback to "prime256v1".
2113 JH/34 SECURITY: Use proper copy of DATA command in error message.
2114 Could leak key material. Remotely exploitable. CVE-2016-9963.
2120 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2121 and 3.4.4 - once the server is enabled to respond to an OCSP request
2122 it does even when not requested, resulting in a stapling non-aware
2123 client dropping the TLS connection.
2125 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2126 support variable-length bit vectors. No functional change.
2128 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2129 The I= interface field on outgoing lines is now after the H= remote
2130 host field, same as incoming lines. There is a separate
2131 outgoing_interface log selector which allows you to disable the
2134 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2135 If not running log_selector +smtp_connection the mainlog would be held
2136 open indefinitely after a "too many connections" event, including to a
2137 deleted file after a log rotate. Leave the per net connection logging
2138 leaving it open for efficiency as that will be quickly detected by the
2139 check on the next write.
2141 HS/01 Bug 1671: Fix post transport crash.
2142 Processing the wait-<transport> messages could crash the delivery
2143 process if the message IDs didn't exist for some reason. When
2144 using 'split_spool_directory=yes' the construction of the spool
2145 file name failed already, exposing the same netto behaviour.
2147 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2148 mime_regex ACL conditions.
2150 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2151 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2152 to HELO, local diagnostic string.
2154 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2155 log line to debug. Even when configured with a tls authenticator many
2156 client connections are expected to not authenticate in this way, so
2157 an authenticate fail is not an error.
2159 HS/02 Add the Exim version string to the process info. This way exiwhat
2160 gives some more detail about the running daemon.
2162 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2163 matter for fast-change records such as DNSBLs.
2165 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2166 retry record, even if constant. There may be multiple transports with
2167 different interface settings and the retry behaviour needs to be kept
2170 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2172 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2174 JH/10 Bug 840: fix log_defer_output option of pipe transport
2176 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2177 hosts_randomize. This matters a lot when combined with mua_wrapper.
2179 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2180 ${quote_pgsql:<string>} operator.
2182 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2184 JH/14 Change header file name for freeradius-client. Relevant if compiling
2185 with Radius support; from the Gentoo tree and checked under Fedora.
2187 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2189 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2190 option transport_filter, meaning no filtering.
2192 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2194 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2195 defaults to "*" (all hosts). The variable is now available when not built
2196 with TLS, default unset, mainly to enable keeping the testsuite sane.
2197 If a server certificate is not supplied (via tls_certificate) an error is
2198 logged, and clients will find TLS connections fail on startup. Presumably
2199 they will retry in-clear.
2200 Packagers of Exim are strongly encouraged to create a server certificate
2201 at installation time.
2203 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2204 with the $config_file variable.
2206 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2207 in transport context, after the attempt, and per-recipient. The latter type
2208 is per host attempted. The event data is the error message, and the errno
2209 information encodes the lookup type (A vs. MX) used for the (first) host,
2210 and the trailing two digits of the smtp 4xx response.
2212 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2213 to write to mainlog (or rejectlog, paniclog) in the window between file
2214 creation and permissions/ownership being changed. Particularly affects
2215 installations where exicyclog is run as root, rather than exim user;
2216 result is that the running daemon panics and dies.
2218 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2220 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2221 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2222 "pri" and "weight". Note that the previous implicit priority given by the
2223 list order is no longer honoured.
2225 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2226 for DKIM processing.
2228 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2229 by defining SUPPORT_SOCKS.
2231 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2232 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2233 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2234 variables are renamed to proxy_{local,external}_{address,port}.
2236 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2237 for a build by defining SUPPORT_I18N
2239 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2240 of the query string, and make ${quote_redis:} do that quoting.
2242 JH/29 Move Events support from Experimental to mainline, enabled by default
2243 and removable for a build by defining DISABLE_EVENT.
2245 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2247 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2248 cached by the daemon.
2250 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2251 by defining LOOKUP_REDIS. The libhiredis library is required.
2253 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2254 keys are given for lookup.
2256 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2257 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2258 only supported when built with TLS support. The PolarSSL SHA routines
2259 are still used when the TLS library is too old for convenient support.
2261 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2262 openssl_options), for security. OpenSSL forces this from version 1.1.0
2263 server-side so match that on older versions.
2265 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2266 allocation for $value could be released as the expansion processing
2267 concluded, but leaving the global pointer active for it.
2269 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2270 and to use the domains and local_parts ACL conditions.
2272 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2273 incorrectly not doubled on cutthrough transmission, hence seen as a
2274 body-termination at the receiving system - resulting in truncated mails.
2275 Commonly the sender saw a TCP-level error, and retransmitted the message
2276 via the normal store-and-forward channel. This could result in duplicates
2277 received - but deduplicating mailstores were liable to retain only the
2278 initial truncated version.
2280 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2282 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2284 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2285 we're in there, support oversigning also; bug 1309.
2287 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2289 HS/04 Add support for keep_environment and add_environment options.
2291 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2292 either intentional arithmetic overflow during PRNG, or testing config-
2295 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2296 delivery resulted in actual delivery. Cancel cutthrough before DATA
2299 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2300 rejects a recipient: pass the reject to the originator.
2302 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2303 Many were false-positives and ignorable, but it's worth fixing the
2306 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2307 for the new environment-manipulation done at startup. Move the routines
2308 from being local to tls.c to being global via the os.c file.
2310 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2311 an extract embedded as result-arg for a map, the first arg for extract
2312 is unavailable so we cannot tell if this is a numbered or keyed
2313 extraction. Accept either.
2319 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2322 JH/02 The smtp transport option "multi_domain" is now expanded.
2324 JH/03 The smtp transport now requests PRDR by default, if the server offers
2327 JH/04 Certificate name checking on server certificates, when exim is a client,
2328 is now done by default. The transport option tls_verify_cert_hostnames
2329 can be used to disable this per-host. The build option
2330 EXPERIMENTAL_CERTNAMES is withdrawn.
2332 JH/05 The value of the tls_verify_certificates smtp transport and main options
2333 default to the word "system" to access the system default CA bundle.
2334 For GnuTLS, only version 3.0.20 or later.
2336 JH/06 Verification of the server certificate for a TLS connection is now tried
2337 (but not required) by default. The verification status is now logged by
2338 default, for both outbound TLS and client-certificate supplying inbound
2341 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2344 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2345 Status Notification (bounce) messages are now MIME format per RFC 3464.
2346 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2347 under the control of the dsn_advertise_hosts option, and routers may
2348 have a dsn_lasthop option.
2350 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2351 default, modifiable by a malware= option. The list separator for
2352 the options can now be changed in the usual way. Bug 68.
2354 JH/10 The smtp_receive_timeout main option is now expanded before use.
2356 JH/11 The incoming_interface log option now also enables logging of the
2357 local interface on delivery outgoing connections.
2359 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2360 if the interface and destination host and port all match.
2362 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2365 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2366 Patch from Andrew Lewis.
2368 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2369 now supports optional time-restrictions, weighting, and priority
2370 modifiers per server. Patch originally by <rommer@active.by>.
2372 JH/16 The spamd_address main option now supports a mixed list of local
2373 and remote servers. Remote servers can be IPv6 addresses, and
2374 specify a port-range.
2376 JH/17 Bug 68: The spamd_address main option now supports an optional
2377 timeout value per server.
2379 JH/18 Bug 1581: Router and transport options headers_add/remove can
2380 now have the list separator specified.
2382 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2385 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2388 JH/21 Support for the A6 type of dns record is withdrawn.
2390 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2391 rather than the verbs used.
2393 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2394 from 255 to 1024 chars.
2396 JH/24 Verification callouts now attempt to use TLS by default.
2398 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2399 are generic router options now. The defaults didn't change.
2401 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2402 Original patch from Alexander Shikoff, worked over by JH.
2404 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2405 files if ZCAT_COMMAND is not executable.
2407 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2409 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2411 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2412 Normally benign, it bites when the pair was led to by a CNAME;
2413 modern usage is to not canonicalize the domain to a CNAME target
2414 (and we were inconsistent anyway for A-only vs AAAA+A).
2416 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2418 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2419 when evaluating $sender_host_dnssec.
2421 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2422 $sender_helo_dnssec variable.
2424 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2426 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2428 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2430 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2431 documented as working, but never had. Support all but $spam_report.
2433 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2434 added for tls authenticator.
2436 HS/03 Add perl_taintmode main config option
2442 TL/01 When running the test suite, the README says that variables such as
2443 no_msglog_check are global and can be placed anywhere in a specific
2444 test's script, however it was observed that placement needed to be near
2445 the beginning for it to behave that way. Changed the runtest perl
2446 script to read through the entire script once to detect and set these
2447 variables, reset to the beginning of the script, and then run through
2448 the script parsing/test process like normal.
2450 TL/02 The BSD's have an arc4random API. One of the functions to induce
2451 adding randomness was arc4random_stir(), but it has been removed in
2452 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2453 function when detected.
2455 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2456 cause callback expansion.
2458 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2459 syntax errors in an expansion can be treated as a string instead of
2460 logging or causing an error, due to the internal use of bool_lax
2461 instead of bool when processing it.
2463 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2464 server certificates when making smtp deliveries.
2466 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2468 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2470 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2472 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2473 Merged patch from Sebastian Wiedenroth.
2475 JH/05 Fix results-pipe from transport process. Several recipients, combined
2476 with certificate use, exposed issues where response data items split
2477 over buffer boundaries were not parsed properly. This eventually
2478 resulted in duplicates being sent. This issue only became common enough
2479 to notice due to the introduction of connection certificate information,
2480 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2482 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2483 size buffer was used, resulting in syntax errors when an expansion
2486 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2487 version 3.3.6 or later.
2489 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2490 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2491 both become "event_action", the variables become $event_name, $event_data
2492 and $event_defer_errno. There is a new variable $verify_mode, usable in
2493 routers, transports and related events. The tls:cert event is now also
2494 raised for inbound connections, if the main configuration event_action
2497 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2498 early OCSP support, but no stapling (appears to be less than 1.0.0).
2500 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2501 server certificate names available under the smtp transport option
2502 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2505 JH/10 Time-related extraction expansions from certificates now use the main
2506 option "timezone" setting for output formatting, and are consistent
2507 between OpenSSL and GnuTLS compilations. Bug 1541.
2509 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2510 encoded parameter in the incoming message. Bug 1558.
2512 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2513 include certificate info, eximon was claiming there were spoolfile
2516 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2518 JH/14 Log delivery-related information more consistently, using the sequence
2519 "H=<name> [<ip>]" wherever possible.
2521 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2522 are problematic for Debian distribution, omit them from the release
2525 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2527 JH/16 Fix string representation of time values on 64bit time_t architectures.
2530 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2531 output list separator was used.
2536 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2537 checkers that were complaining about end of non-void function with no
2540 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2541 This was a regression introduced in 4.83 by another bugfix.
2543 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2545 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2546 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2552 TF/01 Correctly close the server side of TLS when forking for delivery.
2554 When a message was received over SMTP with TLS, Exim failed to clear up
2555 the incoming connection properly after forking off the child process to
2556 deliver the message. In some situations the subsequent outgoing
2557 delivery connection happened to have the same fd number as the incoming
2558 connection previously had. Exim would try to use TLS and fail, logging
2559 a "Bad file descriptor" error.
2561 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2562 utilities have not been installed.
2564 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2565 temporary space as the ACL may create new global variables.
2567 TL/01 LDAP support uses per connection or global context settings, depending
2568 upon the detected version of the libraries at build time.
2570 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2571 to extract and use the src ip:port in logging and expansions as if it
2572 were a direct connection from the outside internet. PPv2 support was
2573 updated based on HAProxy spec change in May 2014.
2575 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2577 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2578 Properly escape header and check for NULL return.
2580 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2583 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2585 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2586 characters in header names, implemented as a verify condition.
2587 Contributed by Michael Fischer v. Mollard.
2589 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2590 results permerror and temperror. Previous values are deprecated but
2591 still accepted. In a future release, err_perm and err_temp will be
2592 completely removed, which will be a backward incompatibility if the
2593 ACL tests for either of these two old results. Patch contributed by
2594 user bes-internal on the mailing list.
2596 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2598 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2599 selectors, in both main and reject logs.
2601 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2604 JH/07 Add malware type "sock" for talking to simple daemon.
2606 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2608 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2609 routers/transports under cutthrough routing.
2611 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2612 numbers. Touch up "bool" conditional to keep the same definition.
2614 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2616 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2618 JH/12 Expand items in router/transport headers_add or headers_remove lists
2619 individually rather than the list as a whole. Bug 1452.
2621 Required for reasonable handling of multiple headers_ options when
2622 they may be empty; requires that headers_remove items with embedded
2623 colons must have them doubled (or the list-separator changed).
2625 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2626 view the policy declared in the DMARC record. Currently, $dmarc_status
2627 is a combined value of both the record presence and the result of the
2630 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2632 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2633 dnslookup router and the smtp transport (applying to the forward
2636 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2637 of ldap servers used for a specific lookup. Patch provided by Heiko
2640 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2641 New variable $lookup_dnssec_authenticated for observability.
2643 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2644 Patch submitted by Lars Timman.
2646 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2648 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2649 Requires trusted mode and valid format message id, aborts otherwise.
2650 Patch contributed by Heiko Schlichting.
2652 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2653 certextract with support for various fields. Bug 1358.
2655 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2656 is requested by default, modifiable by smtp transport option
2659 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2660 operate on certificate variables to give certificate fingerprints
2661 Also new ${sha256:cert_variable}.
2663 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2665 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2666 Christian Aistleitner.
2668 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2670 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2671 file. Patch from Wolfgang Breyha.
2673 JH/25 Expand the coverage of the delivery $host and $host_address to
2674 client authenticators run in verify callout. Bug 1476.
2676 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2677 align with daemon_smtp_ports. Bug 72.
2679 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2680 support and error reporting did not work properly.
2682 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2683 and is readable. Patch from Andrew Colin Kissa.
2685 TL/14 Enhance documentation of ${run expansion and how it parses the
2686 commandline after expansion, particularly in the case when an
2687 unquoted variable expansion results in an empty value.
2689 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2691 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2692 by renaming away from it. Addresses GH issue 15, reported by
2695 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2697 TL/15 SECURITY: prevent double expansion in math comparison functions
2698 (can expand unsanitized data). Not remotely exploitable.
2705 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2707 PP/02 Make -n do something, by making it not do something.
2708 When combined with -bP, the name of an option is not output.
2710 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2713 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2714 $sender_host_name and config options to manage this, and basic check
2717 PP/05 DSCP support for outbound connections and control modifier for inbound.
2719 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2720 (Only plugin which currently uses this is kerberos4, which nobody should
2721 be using, but we should make it available and other future plugins might
2722 conceivably use it, even though it would break NAT; stuff *should* be
2723 using channel bindings instead).
2725 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2726 name; added for Sendmail compatibility; requires admin caller.
2727 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2728 just ignore it); requires trusted caller.
2729 Also parse but ignore: -Ac -Am -X<logfile>
2732 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2734 TL/02 Add +smtp_confirmation as a default logging option.
2736 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2737 Patch by Magnus Holmgren from 2007-02-20.
2739 TL/04 Bugzilla 1281 - Spec typo.
2740 Bugzilla 1283 - Spec typo.
2741 Bugzilla 1290 - Spec grammar fixes.
2743 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2745 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2747 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2748 mailing list by by Dmitry Isaikin.
2750 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2752 JH/02 Support "G" suffix to numbers in ${if comparisons.
2754 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2756 NM/01 Bugzilla 1197 - Spec typo
2757 Bugzilla 1196 - Spec examples corrections
2759 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2761 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2762 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2765 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2766 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2768 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2769 "acl {{name}{arg}...}", and optional args on acl condition
2772 JH/05 Permit multiple router/transport headers_add/remove lines.
2774 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2776 JH/07 Avoid using a waiting database for a single-message-only transport.
2777 Performance patch from Paul Fisher. Bugzilla 1262.
2779 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2782 JH/09 Add $headers_added variable, with content from use of ACL modifier
2783 add_header (but not yet added to the message). Bugzilla 199.
2785 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2786 Pulled from Bugzilla 817 by Wolfgang Breyha.
2788 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2790 (nb: this is the same fix as in Exim 4.80.1)
2792 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2795 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2798 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2800 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2801 advertise SMTP AUTH mechanism to us, instead of a generic
2802 protocol violation error. Also, make Exim more robust to bad
2803 data from the Dovecot auth socket.
2805 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2807 When a queue runner is handling a message, Exim first routes the
2808 recipient addresses, during which it prunes them based on the retry
2809 hints database. After that it attempts to deliver the message to
2810 any remaining recipients. It then updates the hints database using
2813 So if a recipient address works intermittently, it can get repeatedly
2814 deferred at routing time. The retry hints record remains fresh so the
2815 address never reaches the final cutoff time.
2817 This is a fairly common occurrence when a user is bumping up against
2818 their storage quota. Exim had some logic in its local delivery code
2819 to deal with this. However it did not apply to per-recipient defers
2820 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2822 This change adds a proper retry rule check during routing so that the
2823 final cutoff time is checked against the message's age. We only do
2824 this check if there is an address retry record and there is not a
2825 domain retry record; this implies that previous attempts to handle
2826 the address had the retry_use_local_parts option turned on. We use
2827 this as an approximation for the destination being like a local
2828 delivery, as in LMTP.
2830 I suspect this new check makes the old local delivery cutoff check
2831 redundant, but I have not verified this so I left the code in place.
2833 TF/02 Correct gecos expansion when From: is a prefix of the username.
2835 Test 0254 submits a message to Exim with the header
2839 When I ran the test suite under the user fanf2, Exim expanded
2840 the header to contain my full name, whereas it should have added
2841 a Resent-Sender: header. It erroneously treats any prefix of the
2842 username as equal to the username.
2844 This change corrects that bug.
2846 GF/01 DCC debug and logging tidyup
2847 Error conditions log to paniclog rather than rejectlog.
2848 Debug lines prefixed by "DCC: " to remove any ambiguity.
2850 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2852 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2853 Bug spotted by Jeremy Harris; was flawed since initial commit.
2854 Would have resulted in OCSP responses post-SNI triggering an Exim
2855 NULL dereference and crash.
2857 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2859 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2860 Bug detection, analysis and fix by Samuel Thibault.
2861 Bugzilla 1331, Debian bug #698092.
2863 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2865 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2866 Server implementation by Todd Lyons, client by JH.
2867 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2868 config variable "prdr_enable" controls whether the server
2869 advertises the facility. If the client requests PRDR a new
2870 acl_data_smtp_prdr ACL is called once for each recipient, after
2871 the body content is received and before the acl_smtp_data ACL.
2872 The client is controlled by both of: a hosts_try_prdr option
2873 on the smtp transport, and the server advertisement.
2874 Default client logging of deliveries and rejections involving
2875 PRDR are flagged with the string "PRDR".
2877 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2878 fclose(). Diagnosis by Todd Lyons.
2880 PP/17 Update configure.default to handle IPv6 localhost better.
2881 Patch by Alain Williams (plus minor tweaks).
2884 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2885 This is now consistent with GnuTLS, and is now documented: the
2886 previous undocumented portable approach to treating the option as
2887 unset was to force an expansion failure. That still works, and
2888 an empty string is now equivalent.
2890 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2891 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2892 not performing validation itself.
2894 PP/20 Added force_command boolean option to pipe transport.
2895 Patch from Nick Koston, of cPanel Inc.
2897 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2900 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2902 PP/21 Fix eximon continuous updating with timestamped log-files.
2903 Broken in a format-string cleanup in 4.80, missed when I repaired the
2904 other false fix of the same issue.
2905 Report and fix from Heiko Schlichting.
2908 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2909 Report from Prashanth Katuri.
2911 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2912 It's SecureTransport, so affects any MacOS clients which use the
2913 system-integrated TLS libraries, including email clients.
2915 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2916 using a MIME ACL for non-SMTP local injection.
2917 Report and assistance in diagnosis by Warren Baker.
2919 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2921 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2923 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2924 last id that failed so it may be referenced in subsequent ACL's.
2926 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2929 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2930 ldap library initialization, allowing self-signed CA's to be
2931 used. Also properly sets require_cert option later in code by
2932 using NULL (global ldap config) instead of ldap handle (per
2933 session). Bug diagnosis and testing by alxgomz.
2935 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2936 the src/util/ subdirectory.
2938 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2939 renamed to Transport Post Delivery Action by Jeremy Harris, as
2942 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2943 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2944 redis_servers = needs to be configured which will be used by the redis
2945 lookup. Patch from Warren Baker, of The Packet Hub.
2947 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2949 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2950 hostname or reverse DNS when processing a host list. Used suggestions
2951 from multiple comments on this bug.
2953 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2955 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2956 Missed a few lines, added it to make the runtest require no keyboard
2959 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2960 contains upper case chars. Make router use caseful_local_part.
2962 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2963 support when GnuTLS has been built with p11-kit.
2969 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2971 This, or similar/improved, will also be change PP/11 of 4.82.
2977 PP/01 Handle short writes when writing local log-files.
2978 In practice, only affects FreeBSD (8 onwards).
2979 Bugzilla 1053, with thanks to Dmitry Isaikin.
2981 NM/01 Bugzilla 949 - Documentation tweak
2983 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2986 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2988 PP/02 Implemented gsasl authenticator.
2990 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2992 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2993 `pkg-config foo` for cflags/libs.
2995 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2996 with rest of GSASL and with heimdal_gssapi.
2998 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2999 `pkg-config foo` for cflags/libs for the TLS implementation.
3001 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3002 properties get this fed in as external SSF. A number of robustness
3003 and debugging improvements to the cyrus_sasl authenticator.
3005 PP/08 cyrus_sasl server now expands the server_realm option.
3007 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3008 Patch by Jeremy Harris.
3010 PP/10 Added dbmjz lookup type.
3012 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3014 PP/12 MAIL args handles TAB as well as SP, for better interop with
3015 non-compliant senders.
3016 Analysis and variant patch by Todd Lyons.
3018 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3019 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3020 Patch from Dirk Mueller <dmueller@suse.com>
3022 PP/13 tls_peerdn now print-escaped for spool files.
3023 Observed some $tls_peerdn in wild which contained \n, which resulted
3024 in spool file corruption.
3026 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3027 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3028 or write after TLS renegotiation, which otherwise led to messages
3031 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3032 as a tracking header (ie: a signed header comes before the signature).
3033 Patch from Wolfgang Breyha.
3035 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3036 comma-sep list; embedded commas doubled.
3038 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3040 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3042 Report and patch from Dmitry Banschikov.
3044 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3045 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3046 failures. We appear to now support TLS1.1+ with Exim.
3048 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3049 lets Exim select keys and certificates based upon TLS SNI from client.
3050 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3051 before an outbound SMTP session. New log_selector, +tls_sni.
3053 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3054 NULL dereference. Report and patch from Alun Jones.
3056 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3057 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3058 Not seeing resolver debug output on NetBSD, but suspect this is a
3059 resolver implementation change.
3061 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3062 Left warnings. Added "eximon gdb" invocation mode.
3064 PP/21 Defaulting "accept_8bitmime" to true, not false.
3066 PP/22 Added -bw for inetd wait mode support.
3068 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3069 locate the relevant includes and libraries. Made this the default.
3071 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3072 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3074 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3075 This may cause build issues on older platforms.
3077 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3078 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3079 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3080 Added SNI support via GnuTLS too.
3081 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3083 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3085 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3086 Applied second patch from Janne, implementing suggestion to default
3087 multiple-strings-in-record handling to match SPF spec.
3089 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3091 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3092 read-only, out of scope).
3093 Patch from Wolfgang Breyha, report from Stuart Northfield.
3095 PP/29 Fix three issues highlighted by clang analyser static analysis.
3096 Only crash-plausible issue would require the Cambridge-specific
3097 iplookup router and a misconfiguration.
3098 Report from Marcin Mirosław.
3100 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3102 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3103 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3104 As part of this, removing so much warning spew let me fix some minor
3105 real issues in debug logging.
3107 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3108 assignment on my part. Fixed.
3110 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3111 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3112 Janne Snabb (who went above and beyond: thank you).
3114 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3115 string otherwise requires a connection and a bunch more work and it's
3116 relatively easy to get wrong. Should also expose TLS library linkage
3119 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3120 64-bit ${eval} (JH/03).
3122 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3123 GNU libc to support some of the 64-bit stuff, should not lead to
3124 conflicts. Defined before os.h is pulled in, so if a given platform
3125 needs to override this, it can.
3127 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3128 protection layer was required, which is not implemented.
3129 Bugzilla 1254, patch from Wolfgang Breyha.
3131 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3132 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3133 tls_dhparam take prime identifiers. Also unbreak combination of
3134 OpenSSL+DH_params+TLSSNI.
3136 PP/39 Disable SSLv2 by default in OpenSSL support.
3142 PP/01 Solaris build fix for Oracle's LDAP libraries.
3143 Bugzilla 1109, patch from Stephen Usher.
3145 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3147 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3150 TF/02 Fix a couple more cases where we did not log the error message
3151 when unlink() failed. See also change 4.74-TF/03.
3153 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3154 lock up or crash if it happened to be inside a call to libc when it
3155 got a SIGUSR1 from exiwhat.
3157 The SIGUSR1 handler appends the current process status to the process
3158 log which is later printed by exiwhat. It used to use the general
3159 purpose logging code to do this, but several functions it calls are
3160 not safe for signals.
3162 The new output code in the SIGUSR1 handler is specific to the process
3163 log, and simple enough that it's easy to inspect for signal safety.
3164 Removing some special cases also simplifies the general logging code.
3165 Removing the spurious timestamps from the process log simplifies
3168 TF/04 Improved ratelimit ACL condition.
3170 The /noupdate option has been deprecated in favour of /readonly which
3171 has clearer semantics. The /leaky, /strict, and /readonly update modes
3172 are mutually exclusive. The update mode is no longer included in the
3173 database key; it just determines when the database is updated. (This
3174 means that when you upgrade Exim will forget old rate measurements.)
3176 Exim now checks that the per_* options are used with an update mode that
3177 makes sense for the current ACL. For example, when Exim is processing a
3178 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3179 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3180 must specify per_mail/readonly. If you omit the update mode it defaults to
3181 /leaky where that makes sense (as before) or /readonly where required.
3183 The /noupdate option is now undocumented but still supported for
3184 backwards compatibility. It is equivalent to /readonly except that in
3185 ACLs where /readonly is required you may specify /leaky/noupdate or
3186 /strict/noupdate which are treated the same as /readonly.
3188 A useful new feature is the /count= option. This is a generalization
3189 of the per_byte option, so that you can measure the throughput of other
3190 aggregate values. For example, the per_byte option is now equivalent
3191 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3193 The per_rcpt option has been generalized using the /count= mechanism
3194 (though it's more complicated than the per_byte equivalence). When it is
3195 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3196 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3197 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3198 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3199 Note that using per_rcpt with a non-readonly update mode in more than
3200 one ACL will cause the recipients to be double-counted. (The per_mail
3201 and per_byte options don't have this problem.)
3203 The handling of very low rates has changed slightly. If the computed rate
3204 is less than the event's count (usually one) then this event is the first
3205 after a long gap. In this case the rate is set to the same as this event's
3206 count, so that the first message of a spam run is counted properly.
3208 The major new feature is a mechanism for counting the rate of unique
3209 events. The new per_addr option counts the number of different
3210 recipients that someone has sent messages to in the last time period. It
3211 behaves like per_rcpt if all the recipient addresses are different, but
3212 duplicate recipient addresses do not increase the measured rate. Like
3213 the /count= option this is a general mechanism, so the per_addr option
3214 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3215 example, measure the rate that a client uses different sender addresses
3216 with the options per_mail/unique=$sender_address. There are further
3217 details in the main documentation.
3219 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3221 TF/06 Removed a few PCRE remnants.
3223 TF/07 Automatically extract Exim's version number from tags in the git
3224 repository when doing development or release builds.
3226 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3227 Bugzilla 879. Patch from Paul Fisher.
3229 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3230 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3233 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3235 PP/05 Variable $av_failed, true if the AV scanner deferred.
3236 Bugzilla 1078. Patch from John Horne.
3238 PP/06 Stop make process more reliably on build failure.
3239 Bugzilla 1087. Patch from Heiko Schlittermann.
3241 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3242 Bugzilla 1089. Patch from Heiko Schlittermann.
3244 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3245 Bugzilla 1131. Patch from Holger Weiß.
3247 PP/09 Handle IPv6 addresses with SPF.
3248 Bugzilla 860. Patch from Wolfgang Breyha.
3250 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3252 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3255 PP/11 match_* no longer expand right-hand-side by default.
3256 New compile-time build option, EXPAND_LISTMATCH_RHS.
3257 New expansion conditions, "inlist", "inlisti".
3259 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3261 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3263 PP/14 fix log_write() format string regression from TF/03.
3264 Bugzilla 1152. Patch from Dmitry Isaikin.
3270 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3272 PP/02 Harmonised TLS library version reporting; only show if debugging.
3273 Layout now matches that introduced for other libraries in 4.74 PP/03.
3275 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3277 PP/04 New "dns_use_edns0" global option.
3279 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3282 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3283 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3285 TK/01 Updated PolarSSL code to 0.14.2.
3286 Bugzilla 1097. Patch from Andreas Metzler.
3288 PP/07 Catch divide-by-zero in ${eval:...}.
3289 Fixes bugzilla 1102.
3291 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3294 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3295 format-string attack -- SECURITY: remote arbitrary code execution.
3297 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3298 time unintentionally subject to list matching rules, letting the header
3299 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3300 arbitrary string expansion). This allowed for information disclosure.
3302 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3303 INT_MIN/-1 -- value coerced to INT_MAX.
3309 NM/01 Workaround for PCRE version dependency in version reporting
3312 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3313 This fixes portability to compilers other than gcc, notably
3314 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3316 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3317 makefiles for portability to HP-UX and POSIX correctness.
3319 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3320 Also via indented variable definition in the Makefile.
3321 (Debugging by Oliver Heesakkers).
3323 PP/02 Restore caching of spamd results with expanded spamd_address.
3324 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3326 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3327 Improves build reliability. Fix from: Frank Elsner
3329 NM/02 Fix wide character breakage in the rfc2047 coding
3330 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3332 NM/03 Allow underscore in dnslist lookups
3333 Fixes bug 1026. Patch from Graeme Fowler
3335 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3336 Code patches from Adam Ciarcinski of NetBSD.
3338 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3341 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3342 is logged, to avoid truncation. Patch from John Horne.
3344 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3345 Patch from Jakob Hirsch.
3347 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3348 SQL string expansion failure details.
3349 Patch from Andrey Oktyabrski.
3351 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3352 Patch from Simon Arlott.
3354 PP/09 New lookups functionality failed to compile on old gcc which rejects
3355 extern declarations in function scope.
3356 Patch from Oliver Fleischmann
3358 PP/10 Use sig_atomic_t for flags set from signal handlers.
3359 Check getgroups() return and improve debugging.
3360 Fixed developed for diagnosis in bug 927 (which turned out to be
3363 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3364 Patch from Mark Zealey.
3366 PP/12 Bugzilla 1056: Improved spamd server selection.
3367 Patch from Mark Zealey.
3369 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3370 Based on patch from Heiko Schlittermann.
3372 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3373 Patch from Uwe Doering, sign-off by Michael Haardt.
3375 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3376 variable declaration deep within a block. Bug and patch from
3379 PP/15 lookups-Makefile IRIX compatibility coercion.
3381 PP/16 Make DISABLE_DKIM build knob functional.
3383 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3384 Patch by Simon Arlott
3386 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3387 variable argument macros. Our copy now differs from upstream.
3393 TF/01 Failure to get a lock on a hints database can have serious
3394 consequences so log it to the panic log.
3396 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3397 controlled using the smtp_confirmation log selector.
3399 TF/03 Include the error message when we fail to unlink a spool file.
3401 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3402 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3403 for maintaining out-of-tree patches for some time.
3405 PP/01 Bugzilla 139: Documentation and portability issues.
3406 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3407 Handle per-OS dynamic-module compilation flags.
3409 PP/02 Let /dev/null have normal permissions.
3410 The 4.73 fixes were a little too stringent and complained about the
3411 permissions on /dev/null. Exempt it from some checks.
3412 Reported by Andreas M. Kirchwitz.
3414 PP/03 Report version information for many libraries, including
3415 Exim version information for dynamically loaded libraries. Created
3416 version.h, now support a version extension string for distributors
3417 who patch heavily. Dynamic module ABI change.
3419 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3420 privilege escalation vulnerability whereby the Exim run-time user
3421 can cause root to append content of the attacker's choosing to
3424 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3427 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3428 If dropping privileges for untrusted macros, we disabled normal logging
3429 on the basis that it would fail; for the Exim run-time user, this is not
3430 the case, and it resulted in successful deliveries going unlogged.
3431 Fixed. Reported by Andreas Metzler.
3437 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3438 only prepend for the Resent-* case. Fixes regression introduced in
3439 Exim 4.70 by NM/22 for Bugzilla 607.
3441 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3442 increasing numbers of administrators be bitten by this.
3444 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3446 PP/03 Bugzilla 994: added openssl_options main configuration option.
3448 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3450 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3452 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3454 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3455 without a peer certificate, leading to a segfault because of an
3456 assumption that peers always have certificates. Be a little more
3457 paranoid. Problem reported by Martin Tscholak.
3459 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3460 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3461 NB: ClamAV planning to remove STREAM in "middle of 2010".
3462 CL also introduces -bmalware, various -d+acl logging additions and
3463 more caution in buffer sizes.
3465 PP/09 Implemented reverse_ip expansion operator.
3467 PP/10 Bugzilla 937: provide a "debug" ACL control.
3469 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3471 PP/12 Bugzilla 973: Implement --version.
3473 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3475 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3477 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3479 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3480 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3481 ignore trailing whitespace.
3483 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3485 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3488 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3489 Notification from Dr Andrew Aitchison.
3491 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3492 ExtendedDetectionInfo response format.
3493 Notification from John Horne.
3495 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3498 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3499 XSL and documented dependency on system catalogs, with examples of how
3502 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3505 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3506 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3507 configuration file which is writeable by the Exim user or group.
3509 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3510 of configuration files to cover files specified with the -C option if
3511 they are going to be used with root privileges, not just the default
3514 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3515 option (effectively making it always true).
3517 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3518 files to be used while preserving root privileges.
3520 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3521 that rogue child processes cannot use them.
3523 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3524 run-time user, instead of root.
3526 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3527 Exim run-time user without dropping privileges.
3529 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3530 result string, instead of calling string_vformat() twice with the same
3533 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3534 for other users. Others should always drop root privileges if they use
3535 -C on the command line, even for a whitelisted configure file.
3537 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3539 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3545 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3546 $data_path, and $header_path variables; fixed documentation bugs and
3549 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3550 exipick to access non-standard spools, including the "frozen" queue
3553 NM/01 Bugzilla 965: Support mysql stored procedures.
3554 Patch from Alain Williams
3556 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3558 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3559 Patch from Andreas Metzler
3561 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3562 Patch from Kirill Miazine
3564 NM/05 Bugzilla 671: Added umask to procmail example.
3566 JJ/03 installed exipick 20100323.0, fixing doc bug
3568 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3569 directory. Notification and patch from Dan Rosenberg.
3571 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3573 TK/02 Improve log output when DKIM signing operation fails.
3575 MH/01 Treat the transport option dkim_domain as a colon separated
3576 list, not as a single string, and sign the message with each element,
3577 omitting multiple occurences of the same signer.
3579 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3580 Bugzilla 985, 986. Patch by Simon Arlott
3582 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3583 Patch by Simon Arlott
3585 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3586 MBX locking. Notification from Dan Rosenberg.
3592 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3594 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3596 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3598 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3600 NM/04 Bugzilla 915: Flag broken perl installation during build.
3606 TK/01 Added patch by Johannes Berg that expands the main option
3607 "spamd_address" if it starts with a dollar sign.
3609 TK/02 Write list of recipients to X-Envelope-Sender header when building
3610 the mbox-format spool file for content scanning (suggested by Jakob
3613 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3614 (http://www.dcc-servers.net/) support via dccifd. Activated by
3615 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3617 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3618 by Mark Daniel Reidel <mr@df.eu>.
3620 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3621 When building exim an external PCRE library is now needed -
3622 PCRE is a system library on the majority of modern systems.
3623 See entry on PCRE_LIBS in EDITME file.
3625 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3626 conversation. Added nologin parameter to request.
3627 Patch contributed by Kirill Miazine.
3629 TF/01 Do not log submission mode rewrites if they do not change the address.
3631 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3633 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3634 log files in place. Contributed by Roberto Lima.
3636 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3638 TF/03 Bugzilla 615: When checking the local_parts router precondition
3639 after a local_part_suffix or local_part_prefix option, Exim now
3640 does not use the address's named list lookup cache, since this
3641 contains cached lookups for the whole local part.
3643 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3644 Robert Millan. Documentation is in experimental-spec.txt.
3646 TF/04 Bugzilla 668: Fix parallel build (make -j).
3648 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3650 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3651 Patch provided by Jan Srzednicki.
3653 TF/05 Leading white space used to be stripped from $spam_report which
3654 wrecked the formatting. Now it is preserved.
3656 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3657 that they are available at delivery time.
3659 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3661 TF/08 TLS error reporting now respects the incoming_interface and
3662 incoming_port log selectors.
3664 TF/09 Produce a more useful error message if an SMTP transport's hosts
3665 setting expands to an empty string.
3667 NM/06 Bugzilla 744: EXPN did not work under TLS.
3668 Patch provided by Phil Pennock.
3670 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3671 Patch provided by Richard Godbee.
3673 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3674 acl_smtp_notquit, added index entry.
3676 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3677 Patch provided by Eugene Bujak.
3679 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3680 accept(). Patch provided by Maxim Dounin.
3682 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3683 Patch provided by Phil Pennock.
3685 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3687 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3688 Patch provided by Brad "anomie" Jorsch.
3690 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3691 Patch provided by Dean Brooks.
3693 TK/05 Add native DKIM support (does not depend on external libraries).
3695 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3696 Patch provided by Graeme Fowler.
3698 NM/16 Bugzilla 851: Documentation example syntax fix.
3700 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3702 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3705 NM/19 Bugzilla 745: TLS version reporting.
3706 Patch provided by Phil Pennock.
3708 NM/20 Bugzilla 167: bool: condition support.
3709 Patch provided by Phil Pennock.
3711 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3712 clients. Patch provided by Phil Pennock.
3714 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3715 Patch provided by Brad "anomie" Jorsch.
3717 NM/23 Bugzilla 687: Fix misparses in eximstats.
3718 Patch provided by Heiko Schlittermann.
3720 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3721 Patch provided by Heiko Schlittermann.
3723 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3724 plus update to original patch.
3726 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3728 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3729 Patch provided by David Brownlee.
3731 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3733 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3735 NM/30 Bugzilla 888: TLS documentation bugfixes.
3737 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3739 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3740 Unlike the original bugzilla I have changed all shell scripts in src tree.
3742 NM/33 Bugzilla 898: Transport filter timeout fix.
3743 Patch by Todd Rinaldo.
3745 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3746 Patch by Serge Demonchaux.
3748 NM/35 Bugzilla 39: Base64 decode bug fixes.
3749 Patch by Jakob Hirsch.
3751 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3753 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3755 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3757 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3763 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3764 ALT-N's libdkim that I have put here:
3765 http://duncanthrax.net/exim-experimental/
3767 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3768 were called 'true' and it seems that C99 defines that as a reserved
3769 keyword to be used with 'bool' variable types. That means you could
3770 not include C99-style headers which use bools without triggering
3771 build errors in sieve.c.
3773 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3774 as mailq or other aliases. Changed the --help handling significantly
3775 to do whats expected. exim_usage() emits usage/help information.
3777 SC/01 Added the -bylocaldomain option to eximstats.
3779 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3781 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3783 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3789 PH/01 Another patch from the Sieve maintainer.
3791 PH/02 When an IPv6 address is converted to a string for single-key lookup
3792 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3793 dots are used instead of colons so that keys in lsearch files need not
3794 contain colons. This was done some time before quoting was made available
3795 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3796 (notated using the quote facility) so as to distinguish them from IPv4
3797 keys. This meant that lookups for IP addresses in host lists did not work
3798 for iplsearch lookups.
3800 This has been fixed by arranging for IPv6 addresses to be expressed with
3801 colons if the lookup type is iplsearch. This is not incompatible, because
3802 previously such lookups could never work.
3804 The situation is now rather anomalous, since one *can* have colons in
3805 ordinary lsearch keys. However, making the change in all cases is
3806 incompatible and would probably break a number of configurations.
3808 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3811 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3812 conversion specification without a maximum field width, thereby enabling
3813 a rogue spamd server to cause a buffer overflow. While nobody in their
3814 right mind would setup Exim to query an untrusted spamd server, an
3815 attacker that gains access to a server running spamd could potentially
3816 exploit this vulnerability to run arbitrary code as the Exim user.
3818 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3819 $primary_hostname instead of what libspf2 thinks the hosts name is.
3821 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3822 a directory entry by the name of the lookup key. Previously, if a
3823 symlink pointed to a non-existing file or a file in a directory that
3824 Exim lacked permissions to read, a lookup for a key matching that
3825 symlink would fail. Now it is enough that a matching directory entry
3826 exists, symlink or not. (Bugzilla 503.)
3828 PH/03 The body_linecount and body_zerocount variables are now exported in the
3831 PH/04 Added the $dnslist_matched variable.
3833 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3834 This means they are set thereafter only if the connection becomes
3837 PH/06 Added the client_condition to authenticators so that some can be skipped
3838 by clients under certain conditions.
3840 PH/07 The error message for a badly-placed control=no_multiline_responses left
3841 "_responses" off the end of the name.
3843 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3845 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3846 (without spaces) instead of just copying the configuration text.
3848 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3850 PH/11 Added $max_received_linelength.
3852 PH/12 Added +ignore_defer and +include_defer to host lists.
3854 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3855 way in which PCRE > 7.0 is built.
3857 PH/14 Implemented queue_only_load_latch.
3859 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3860 MAIL command. The effect was to mangle the value on 64-bit systems.
3862 PH/16 Another patch from the Sieve maintainer.
3864 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3866 PH/18 If a system quota error occurred while trying to create the file for
3867 a maildir delivery, the message "Mailbox is full" was not appended to the
3868 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3869 only to a quota excession during the actual writing of the file.
3871 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3872 characters?) which causes problems in log lines. The DN values are now
3873 passed through string_printing() before being added to log lines.
3875 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3876 and InterBase are left for another time.)
3878 PH/21 Added message_body_newlines option.
3880 PH/22 Guard against possible overflow in moan_check_errorcopy().
3882 PH/23 POSIX allows open() to be a macro; guard against that.
3884 PH/24 If the recipient of an error message contained an @ in the local part
3885 (suitably quoted, of course), incorrect values were put in $domain and
3886 $local_part during the evaluation of errors_copy.
3892 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3893 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3896 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3897 issue a MAIL command.
3899 PH/02 In an ACL statement such as
3901 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3903 if a client was not listed at all, or was listed with a value other than
3904 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3905 the condition was not true (as it should be), so access was not denied.
3906 The bug was that the ! inversion was incorrectly passed on to the second
3907 item. This has been fixed.
3909 PH/03 Added additional dnslists conditions == and =& which are different from
3910 = and & when the dns lookup returns more than one IP address.
3912 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3913 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3915 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3916 FSYNC, which compiles an option called disable_fsync that allows for
3917 bypassing fsync(). The documentation is heavily laced with warnings.
3919 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3921 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3922 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3923 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3924 including adding "make clean"; (3) Added -fPIC when compiling the test
3925 dynamically loaded module, to get rid of a warning.
3927 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3928 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3929 = 0s. The bug is otherwise harmless.
3931 PH/07 There was a bug in the dovecot authenticator such that the value of
3932 $auth1 could be overwritten, and so not correctly preserved, after a
3933 successful authentication. This usually meant that the value preserved by
3934 the server_setid option was incorrect.
3936 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3938 PH/09 Installed PCRE release 7.0.
3940 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3941 run for batched SMTP input. It is now run at the start of every message
3942 in the batch. While fixing this I discovered that the process information
3943 (output by running exiwhat) was not always getting set for -bs and -bS
3944 input. This is fixed, and it now also says "batched" for BSMTP.
3946 PH/11 Added control=no_pipelining.
3948 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3949 patch, slightly modified), and move the expansion of helo_data till after
3950 the connection is made in the smtp transport (so it can use these
3953 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3955 PH/14 Added log_selector = +pid.
3957 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3959 PH/16 Add ${if forany and ${if forall.
3961 PH/17 Added dsn_from option to vary the From: line in DSNs.
3963 PH/18 Flush SMTP output before performing a callout, unless control =
3964 no_callout_flush is set.
3966 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3967 was true (the default) a successful delivery failed to delete the retry
3968 item, thus causing premature timeout of the address. The bug is now
3971 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3973 PH/21 Long custom messages for fakedefer and fakereject are now split up
3974 into multiline responses in the same way that messages for "deny" and
3975 other ACL rejections are.
3977 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3978 with slight modification.
3980 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3981 draft, changing the syntax and factoring some duplicate code".
3983 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3984 for deliveries of the second and subsequent messages over the same SMTP
3987 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3988 ${reduce, with only minor "tidies".
3990 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3992 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3993 expansion side effects.
3995 PH/27 When a message times out after an over-quota error from an Exim-imposed
3996 quota, the bounce message says "mailbox is full". This message was not
3997 being given when it was a system quota that was exceeded. It now should
4000 MH/03 Made $recipients available in local_scan(). local_scan() already has
4001 better access to the recipient list through recipients_list[], but
4002 $recipients can be useful in postmaster-provided expansion strings.
4004 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4005 in the case of a MAIL command with additional options following the
4006 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4007 were accidentally chopped off.
4009 PH/29 SMTP synchronization checks are implemented when a command is read -
4010 there is a check that no more input is waiting when there shouldn't be
4011 any. However, for some commands, a delay in an ACL can mean that it is
4012 some time before the response is written. In this time, more input might
4013 arrive, invalidly. So now there are extra checks after an ACL has run for
4014 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4015 pipelining has not been advertised.
4017 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4019 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4020 returned in the SMTP response when smtp_return_error_details was false.
4021 This has been fixed.
4023 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4024 instead of the C I/O that was originally supplied, because problems were
4025 reported on Solaris.
4027 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4028 Exim which did not show up earlier: it was assuming that a call to
4029 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4030 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4031 was a macro that became an assignment, so it seemed to work. This has
4032 changed to a proper function call with a void return, hence the compile
4033 error. Exim's code has been fixed.
4035 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4038 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4040 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4042 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4043 because it turns out to be too big or there is a timeout) there is no
4044 "Completed" line in the log. When some messages of this type were
4045 selected by exigrep, they were listed as "not completed". Others were
4046 picked up by some special patterns. I have improved the selection
4047 criteria to be more general.
4049 PH/38 The host_find_failed option in the manualroute router can now be set
4050 to "ignore", to completely ignore a host whose IP address cannot be
4051 found. If all hosts are ignored, the behaviour is controlled by the new
4052 host_all_ignored option.
4054 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4055 homing or because of multiple MX records with /mx) generated more than
4056 one IP address, and the following item turned out to be the local host,
4057 all the secondary addresses of the first item were incorrectly removed
4058 from the list, along with the local host and any following hosts (which
4059 is what is supposed to happen).
4061 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4062 whoever called Exim into the -H file. In the case of the daemon it was
4063 behaving confusingly. When first started, it used values for whoever
4064 started the daemon, but after a SIGHUP it used the Exim user (because it
4065 calls itself on a restart). I have changed the code so that it now always
4068 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4069 message are rejected with the same error (e.g. no authentication or bad
4070 sender address), and a DATA command is nevertheless sent (as can happen
4071 with PIPELINING or a stupid MUA), the error message that was given to the
4072 RCPT commands is included in the rejection of the DATA command. This is
4073 intended to be helpful for MUAs that show only the final error to their
4076 PH/42 Another patch from the Sieve maintainer.
4078 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4079 Eximstats - Fixed some broken HTML links and added missing column headers
4081 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4082 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4084 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4086 PH/43 Yet another patch from the Sieve maintainer.
4088 PH/44 I found a way to check for a TCP/IP connection going away before sending
4089 the response to the final '.' that terminates a message, but only in the
4090 case where the client has not sent further data following the '.'
4091 (unfortunately, this is allowed). However, in many cases there won't be
4092 any further data because there won't be any more messages to send. A call
4093 to select() can be used: if it shows that the input is "ready", there is
4094 either input waiting, or the socket has been closed. An attempt to read
4095 the next input character can distinguish the two cases. Previously, Exim
4096 would have sent an OK response which the client would never have see.
4097 This could lead to message repetition. This fix should cure that, at
4098 least in a lot of common cases.
4100 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4101 advertised in response to EHLO.
4107 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4108 fixed by 4.65/MH/01 (is this a record?) are fixed:
4110 (i) An empty string was always treated as zero by the numeric comparison
4111 operators. This behaviour has been restored.
4113 (ii) It is documented that the numeric comparison operators always treat
4114 their arguments as decimal numbers. This was broken in that numbers
4115 starting with 0 were being interpreted as octal.
4117 While fixing these problems I realized that there was another issue that
4118 hadn't been noticed. Values of message_size_limit (both the global option
4119 and the transport option) were treated as octal if they started with 0.
4120 The documentation was vague. These values are now always treated as
4121 decimal, and I will make that clear in the documentation.
4127 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4128 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4131 MH/01 Don't check that the operands of numeric comparison operators are
4132 integers when their expansion is in "skipping" mode (fixes bug
4133 introduced by 4.64-PH/07).
4135 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4136 child addresses, Exim now panics and dies. Previously, because the count
4137 is held in a short int, deliveries were likely to be lost. As such a
4138 large number of recipients for a single message is ridiculous
4139 (performance will be very, very poor), I have chosen to impose a limit
4140 rather than extend the field.
4146 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4147 leftover -K file (the existence of which was triggered by #402).
4148 While we were at it, introduced process PID as part of the -K
4149 filename. This should rule out race conditions when creating
4152 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4153 processing considerably. Previous code took too long for large mails,
4154 triggering a timeout which in turn triggers #401.
4156 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4157 in the DK code in transports.c. sendfile() is not really portable,
4158 hence the _LINUX specificness.
4160 TF/01 In the add_headers option to the mail command in an Exim filter,
4161 there was a bug that Exim would claim a syntax error in any
4162 header after the first one which had an odd number of characters
4165 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4166 callout verification, Exim cached a "reject" for the entire domain. This
4167 is correct for most verifications, but it is not correct for a recipient
4168 verification with use_sender or use_postmaster set, because in that case
4169 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4170 case of MAIL FROM:<> rejection from other early rejections (e.g.
4171 rejection of HELO). When verifying a recipient using a non-null MAIL
4172 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4173 Whatever the result of the callout, the value of the domain cache is
4174 left unchanged (for any other kind of callout, getting as far as trying
4175 RCPT means that the domain itself is ok).
4177 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4180 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4181 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4182 session, the daemon ignores ECONNECT errors and logs others; it now
4183 ignores EPIPE as well.
4185 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4186 (quoted-printable decoding).
4188 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4189 later the small subsequent patch to fix an introduced bug.
4191 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4193 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4195 PH/08 An error is now given if message_size_limit is specified negative.
4197 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4198 to be given (somewhat) arbitrary names.
4200 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4203 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4204 miscellaneous code fixes
4206 PH/10 Added the log_reject_target ACL modifier to specify where to log
4209 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4210 hostname. This is wrong, because it relates to the incoming message (and
4211 probably the interface on which it is arriving) and not to the outgoing
4212 callout (which could be using a different interface). This has been
4213 changed to use the value of the helo_data option from the smtp transport
4214 instead - this is what is used when a message is actually being sent. If
4215 there is no remote transport (possible with a router that sets up host
4216 addresses), $smtp_active_hostname is used.
4218 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4219 tweaks were necessary in order to get it to work (see also 21 below):
4220 (a) The code assumed that strncpy() returns a negative number on buffer
4221 overflow, which isn't the case. Replaced with Exim's string_format()
4223 (b) There were several signed/unsigned issues. I just did the minimum
4224 hacking in of casts. There is scope for a larger refactoring.
4225 (c) The code used strcasecmp() which is not a standard C function.
4226 Replaced with Exim's strcmpic() function.
4227 (d) The code set only $1; it now sets $auth1 as well.
4228 (e) A simple test gave the error "authentication client didn't specify
4229 service in request". It would seem that Dovecot has changed its
4230 interface. Fortunately there's a specification; I followed it and
4231 changed what the client sends and it appears to be working now.
4233 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4236 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4237 address A is aliased to B and C, where B exists and C does not. Without
4238 -v the output is "A verified" because verification stops after a
4239 successful redirection if more than one address is generated. However,
4240 with -v the child addresses are also verified. Exim was outputting "A
4241 failed to verify" and then showing the successful verification for C,
4242 with its parentage. It now outputs "B failed to verify", showing B's
4243 parentage before showing the successful verification of C.
4245 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4246 look up a TXT record in a specific list after matching in a combined
4249 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4250 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4251 they consult the DNS. I had assumed they would set it the way they
4252 wanted; and indeed my experiments on Linux seem to show that in some
4253 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4254 To be on the safe side, however, I have now made the interface to
4255 host_find_byname() similar to host_find_bydns(), with an argument
4256 containing the DNS resolver options. The host_find_byname() function now
4257 sets these options at its start, just as host_find_bydns() does. The smtp
4258 transport options dns_qualify_single and dns_search_parents are passed to
4259 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4260 of host_find_byname() use the default settings of RES_DEFNAMES
4261 (qualify_single) but not RES_DNSRCH (search_parents).
4263 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4264 spool_read_header() do less string testing, by means of a preliminary
4265 switch on the second character of optional "-foo" lines. (This is
4266 overdue, caused by the large number of possibilities that now exist.
4267 Originally there were few.) While I was there, I also converted the
4268 str(n)cmp tests so they don't re-test the leading "-" and the first
4269 character, in the hope this might squeeze out yet more improvement.
4271 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4272 flag allowing group syntax was set by the header_syntax check but not
4273 turned off, possible causing trouble later; (2) The flag was not being
4274 set at all for the header_verify test, causing "group"-style headers to
4275 be rejected. I have now set it in this case, and also caused header_
4276 verify to ignore an empty address taken from a group. While doing this, I
4277 came across some other cases where the code for allowing group syntax
4278 while scanning a header line wasn't quite right (mostly, not resetting
4279 the flag correctly in the right place). These bugs could have caused
4280 trouble for malformed header lines. I hope it is now all correct.
4282 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4283 with the "reply" argument non-NULL. The code, however (which originally
4284 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4285 but it didn't always do it. This confused somebody who was copying the
4286 code for some other use. I have removed all the tests.
4288 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4289 feature that was used to support insecure browsers during the U.S. crypto
4290 embargo. It requires special client support, and Exim is probably the
4291 only MTA that supported it -- and would never use it because real RSA is
4292 always available. This code has been removed, because it had the bad
4293 effect of slowing Exim down by computing (never used) parameters for the
4294 RSA_EXPORT functionality.
4296 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4297 authenticator to fail if there's a tab character in the incoming data
4298 (there should never be unless someone is messing about, as it's supposed
4299 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4300 if the connection is using TLS or if the remote IP is the same as the
4301 local IP, and the "valid-client-cert option" if a client certificate has
4304 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4305 authenticators. This can be used for authorization after authentication
4306 succeeds. (In the case of plaintext, it servers for both authentication
4309 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4310 if any retry times were supplied.
4312 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4313 connection, where there is no client IP address to check. In this
4314 situation, the verify now always succeeds.
4316 PH/25 Applied John Jetmore's -Mset patch.
4318 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4320 PH/27 In a string expansion for a processed (not raw) header when multiple
4321 headers of the same name were present, leading whitespace was being
4322 removed from all of them, but trailing whitespace was being removed only
4323 from the last one. Now trailing whitespace is removed from each header
4324 before concatenation. Completely empty headers in a concatenation (as
4325 before) are ignored.
4327 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4328 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4330 PH/29 [Removed. This was a change that I later backed out, and forgot to
4331 correct the ChangeLog entry (that I had efficiently created) before
4332 committing the later change.]
4334 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4335 address errors (4xx response to RCPT) over the same connection as other
4336 messages routed to the same hosts. Such deliveries are always "forced",
4337 so retry times are not inspected. This resulted in far too many retries
4338 for the affected addresses. The effect occurred only when there were more
4339 hosts than the hosts_max_try setting in the smtp transport when it had
4340 the 4xx errors. Those hosts that it had tried were not added to the list
4341 of hosts for which the message was waiting, so if all were tried, there
4342 was no problem. Two fixes have been applied:
4344 (i) If there are any address or message errors in an SMTP delivery, none
4345 of the hosts (tried or untried) are now added to the list of hosts
4346 for which the message is waiting, so the message should not be a
4347 candidate for sending over the same connection that was used for a
4348 successful delivery of some other message. This seems entirely
4349 reasonable: after all the message is NOT "waiting for some host".
4350 This is so "obvious" that I'm not sure why it wasn't done
4351 previously. Hope I haven't missed anything, but it can't do any
4352 harm, as the worst effect is to miss an optimization.
4354 (ii) If, despite (i), such a delivery is accidentally attempted, the
4355 routing retry time is respected, so at least it doesn't keep
4356 hammering the server.
4358 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4359 in ${readsocket because some servers need this prod.
4361 PH/32 Added some extra debug output when updating a wait-xxx database.
4363 PH/33 The hint "could be header name not terminated by colon", which has been
4364 given for certain expansion errors for a long time, was not being given
4365 for the ${if def:h_colon_omitted{... case.
4367 PH/34 The spec says: "With one important exception, whenever a domain list is
4368 being scanned, $domain contains the subject domain." There was at least
4369 one case where this was not true.
4371 PH/35 The error "getsockname() failed: connection reset by peer" was being
4372 written to the panic log as well as the main log, but it isn't really
4373 panic-worthy as it just means the connection died rather early on. I have
4374 removed the panic log writing for the ECONNRESET error when getsockname()
4377 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4378 runs only) independently of the message's sender address. This meant
4379 that, if the 4xx error was in fact related to the sender, a different
4380 message to the same recipient with a different sender could confuse
4381 things. In particular, this can happen when sending to a greylisting
4382 server, but other circumstances could also provoke similar problems.
4383 I have changed the default so that the retry time for these errors is now
4384 based a combination of the sender and recipient addresses. This change
4385 can be overridden by setting address_retry_include_sender=false in the
4388 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4389 remote server are returned as part of bounce messages. This was not
4390 happening for LMTP over a pipe (the lmtp transport), but now it is the
4391 same for both kinds of LMTP.
4393 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4394 in header lines that were in fact CNAMEs. This is no longer the case.
4396 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4397 and queue runs started by the daemon processed all messages. This has
4398 been fixed so that -R and -S can now usefully be given with -q<time>.
4400 PH/40 Import PCRE release 6.7 (fixes some bugs).
4402 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4404 PH/42 Give an error if -q is specified more than once.
4406 PH/43 Renamed the variables $interface_address and $interface_port as
4407 $received_ip_address and $received_port, to make it clear that these
4408 values apply to message reception, and not to the outgoing interface when
4409 a message is delivered. (The old names remain recognized, of course.)
4411 PH/44 There was no timeout on the connect() call when using a Unix domain
4412 socket in the ${readsocket expansion. There now is.
4414 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4415 be meaningful with "accept".
4417 SC/01 Eximstats V1.43
4418 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4420 SC/02 Eximstats V1.44
4421 Use a glob alias rather than an array ref in the generated
4422 parser. This improves both readability and performance.
4424 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4425 Collect SpamAssassin and rejection statistics.
4426 Don't display local sender or destination tables unless
4427 there is data to show.
4428 Added average volumes into the top table text output.
4430 SC/04 Eximstats V1.46
4431 Collect data on the number of addresses (recipients)
4432 as well as the number of messages.
4434 SC/05 Eximstats V1.47
4435 Added 'Message too big' to the list of mail rejection
4436 reasons (thanks to Marco Gaiarin).
4438 SC/06 Eximstats V1.48
4439 Mainlog lines which have GMT offsets and are too short to
4440 have a flag are now skipped.
4442 SC/07 Eximstats V1.49 (Alain Williams)
4443 Added the -emptyok flag.
4445 SC/08 Eximstats V1.50
4446 Fixes for obtaining the IP address from reject messages.
4448 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4449 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4450 whitespace changes from 4.64-PH/27
4452 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4455 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4456 are found, allow negative numbers in numeric criteria)
4458 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4460 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4461 to match changes made in 4.64-PH/43
4463 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4465 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4466 "domain" is actually an IP address, and force a failure. This locks out
4467 those revolvers/nameservers that support "A-for-A" lookups, in
4468 contravention of the specifications.
4470 PH/48 When a host name was looked up from an IP address, and the subsequent
4471 forward lookup of the name timed out, the host name was left in
4472 $sender_host_name, contrary to the specification.
4474 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4475 restricted to single-key lookups, Exim was not diagnosing an error if
4476 * or *@ was used with a query-style lookup.
4478 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4480 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4481 long ago, but noone interested enough thought of it. Let's just say that
4482 the "1.1" means that there are some new functions that weren't there at
4483 some point in the past.
4485 PH/51 Error processing for expansion failure of helo_data from an smtp
4486 transport during callout processing was broken.
4488 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4489 tested/used via the -bh/-bhc/-bs options.
4491 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4492 bug, fixed in subsequent PCRE releases).
4494 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4495 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4497 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4503 SC/01 Use a glob alias rather than an array ref in eximstats generated
4504 parser. This improves both readability and performance.
4506 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4507 Don't display local sender or destination tables in eximstats unless
4508 there is data to show.
4509 Added average volumes into the eximstats top table text output.
4511 SC/03 Collect data on the number of addresses (recipients) as well
4512 as the number of messages in eximstats.
4514 TF/01 Correct an error in the documentation for the redirect router. Exim
4515 does (usually) call initgroups() when daemonizing.
4517 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4518 with consistent privilege compared to when running as a daemon.
4520 TF/03 Note in the spec that $authenticated_id is not set for local
4521 submissions from trusted users.
4523 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4524 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4526 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4527 by adding some example configuration directives to the default
4528 configuration file. A little bit of work is required to uncomment the
4529 directives and define how usernames and passwords are checked, but
4530 there is now a framework to start from.
4532 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4533 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4534 without this. I don't know how relevant this is to other LDAP libraries.
4536 PH/02 Add the verb name to the "unknown ACL verb" error.
4538 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4540 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4542 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4543 directory not expanded when it should be if an expanded home directory
4544 was set for the address (which is overridden by the transport).
4546 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4549 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4550 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4551 because it is too late at that time, and has no effect.
4553 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4554 security issue with \' (bugzilla #107). I could not use the
4555 PQescapeStringConn() function, because it needs a PGconn value as one of
4558 PH/08 When testing addresses using -bt, indicate those final addresses that
4559 are duplicates that would not cause an additional delivery. At least one
4560 person was confused, thinking that -bt output corresponded to deliveries.
4561 (Suppressing duplicates isn't a good idea as you lose the information
4562 about possibly different redirections that led to the duplicates.)
4564 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4565 systems where poll() doesn't work, in particular OS X.
4567 PH/10 Added more information to debugging output for retry time not reached.
4569 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4570 operations in malware.c.
4572 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4575 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4576 syslog_duplication set false, log lines that would normally be written
4577 both the the main log and to the reject log were not written to syslog at
4580 PH/14 In the default configuration, change the use of "message" in ACL warn
4581 statements to "add_header".
4583 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4584 not followed by a command (e.g. "seen endif").
4586 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4587 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4590 PH/17 Added extra conditions to the default value of delay_warning_condition
4594 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4595 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4596 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4599 The Auto-Submitted: and various List- headers are standardised, whereas I
4600 don't think Precedence: ever was.
4602 PH/18 Refactored debugging code in route_finduser() to show more information,
4603 in particular, the error code if getpwnam() issues one.
4605 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4606 This is apparently needed in addition to the PH/07 change above to avoid
4607 any possible encoding problems.
4609 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4610 but not after initializing Perl.
4612 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4613 output them only if debugging. By default they are written stderr,
4614 apparently, which is not desirable.
4616 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4619 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4622 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4624 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4625 authenticated or an ident call has been made. Suppress the default
4626 values for $authenticated_id and $authenticated_sender (but permit -oMai
4627 and -oMas) when testing with -bh.
4629 PH/24 Re-jigged the order of the tests in the default configuration so that the
4630 tests for valid domains and recipients precede the DNS black list and CSA
4631 tests, on the grounds that those ones are more expensive.
4633 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4634 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4635 This bug exists in every version of Exim that I still have, right back to
4638 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4639 However, an attempt to turn on case-sensitivity in a regex key by
4640 including (?-i) didn't work because the subject string was already
4641 lowercased, and the effects were non-intuitive. It turns out that a
4642 one-line patch can be used to allow (?-i) to work as expected.
4648 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4649 other effects) broke the use of negated acl sub-conditions.
4651 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4654 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4655 "Deny" causes Exim to reject the incoming connection with a 554 error.
4656 Unfortunately, if there is a major crisis, such as a disk failure,
4657 tcp-wrappers gives "deny", whereas what one would like would be some
4658 kind of temporary error. A kludge has been added to help with this.
4659 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4660 554 error is used if errno is still zero or contains ENOENT (which occurs
4661 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4664 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4666 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4667 errors. Otherwise a message that provokes a temporary error (when other
4668 messages do not) can cause a whole host to time out.
4670 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4671 addresses were routed directly to files or pipes from a redirect router.
4672 File deliveries just didn't batch; pipe deliveries might have suffered
4675 PH/06 A failure to get a lock for a hints database would erroneously always say
4676 "Failed to get write lock", even when it was really a read lock.
4678 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4679 of 0600. This has been changed to use the value of the lockfile_mode
4680 option (which defaults to 0600).
4682 PH/08 Applied small patch from the Sieve maintainer.
4684 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4685 folder from quota calculations, a direct delivery into this folder messed
4686 up the contents of the maildirsize file. This was because the regex was
4687 used only to exclude .Trash (or whatever) when the size of the mailbox
4688 was calculated. There was no check that a delivery was happening into an
4689 excluded directory. This bug has been fixed by ignoring all quota
4690 processing for deliveries into excluded directories.
4692 PH/10 Added the maildirfolder_create_regex option to appendfile.
4698 PH/01 The code for finding all the local interface addresses on a FreeBSD
4699 system running IPv6 was broken. This may well have applied to all BSD
4700 systems, as well as to others that have similar system calls. The broken
4701 code found IPv4 interfaces correctly, but gave incorrect values for the
4702 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4703 that it would not match correctly against @[] and not recognize the IPv6
4706 PH/02 The ipliteral router was not recognizing addresses of the form user@
4707 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4709 PH/03 Added disable_ipv6.
4711 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4712 decoded form, because it is most often used to construct To: headers
4713 lines in autoreplies, and the decoded form may well be syntactically
4714 invalid. However, $reply_address has leading white space removed, and all
4715 newlines turned into spaces so that the autoreply transport does not
4718 PH/05 If group was specified without a user on a router, and no group or user
4719 was specified on a transport, the group from the router was ignored.
4721 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4722 for visible compile-time settings that can be used to change these
4723 numbers, for those that want even more. Backwards compatibility with old
4724 spool files has been maintained. However, going back to a previous Exim
4725 release will lost any variables that are in spool files.
4727 PH/07 Two small changes when running in the test harness: increase delay when
4728 passing a TCP/IP connection to a new process, in case the original
4729 process has to generate a bounce, and remove special handling of
4730 127.0.0.2 (sic), which is no longer necessary.
4732 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4733 be the same on different OS.
4735 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4738 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4739 whether --show-vars was specified or not
4741 JJ/02 exipick: Added support for new ACL variable spool format introduced
4744 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4745 syntactically invalid From: or Reply-to: line, and a filter used this to
4746 generate an autoreply, and therefore failed to obtain an address for the
4747 autoreply, Exim could try to deliver to a non-existent relative file
4748 name, causing unrelated and misleading errors. What now happens is that
4749 it logs this as a hard delivery error, but does not attempt to create a
4752 PH/11 The exinext utility has a -C option for testing purposes, but although
4753 the given file was scanned by exinext itself; it wasn't being passed on
4754 when Exim was called.
4756 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4757 an end-of-file indication when reading a command response.
4759 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4760 compiled. In many other places in Exim, IPv6 addresses are always
4761 recognized, so I have changed this. It also means that IPv4 domain
4762 literals of the form [IPV4:n.n.n.n] are now always recognized.
4764 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4765 used if the router is not running as root, for example, when verifying at
4766 ACL time, or when using -bh. The debugging output from this situation was
4767 non-existent - all you got was a failure to exec. I have made two
4770 (a) Failures to set uid/gid, the current directory, or a process leader
4771 in a subprocess such as that created by queryprogram now generate
4772 suitable debugging output when -d is set.
4774 (b) The queryprogram router detects when it is not running as root,
4775 outputs suitable debugging information if -d is set, and then runs
4776 the subprocess without attempting to change uid/gid.
4778 PH/15 Minor change to Makefile for building test_host (undocumented testing
4781 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4782 additional section of a DNS packet that returns MX or SRV records.
4783 Instead, it always explicitly searches for A/AAAA records. This avoids
4784 major problems that occur when a DNS server includes only records of one
4785 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4786 fixed another bug: if SRV records were looked up and the corresponding
4787 address records were *not* found in the additional section, the port
4788 values from the SRV records were lost.
4790 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4791 using the correct key (the original address) when searching the retry
4792 rules in order to find which one to use for generating the retry hint.
4794 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4795 adding the default one. Similarly, if it contains a Reply-To: header, the
4796 errors_reply_to option, if set, is not used.
4798 PH/19 When calculating a retry time, Exim used to measure the "time since
4799 failure" by looking at the "first failed" field in the retry record. Now
4800 it does not use this if it is later than than the arrival time of the
4801 message. Instead it uses the arrival time. This makes for better
4802 behaviour in cases where some deliveries succeed, thus re-setting the
4803 "first failed" field. An example is a quota failure for a huge message
4804 when small messages continue to be delivered. Without this change, the
4805 "time since failure" will always be short, possible causing more frequent
4806 delivery attempts for the huge message than are intended.
4807 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4809 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4810 $1, $2, $3) because the numerical variables can be reset during some
4811 expansion items (e.g. "match"), thereby losing the authentication data.
4813 PH/21 Make -bV show the size of off_t variables so that the test suite can
4814 decide whether to run tests for quotas > 2G.
4816 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4817 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4818 filecount value is greater than 2G or if a quota value is greater than 2G
4819 on a system where the size of off_t is not greater than 4, a panic error
4822 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4823 never match. The debug and -bh output now contains an explicit error
4824 message indicating a malformed IPv4 address or mask.
4826 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4827 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4828 PH/23 above applies.
4830 PH/25 Do not write to syslog when running in the test harness. The only
4831 occasion when this arises is a failure to open the main or panic logs
4832 (for which there is an explicit test).
4834 PH/26 Added the /no_tell option to "control=freeze".
4836 PH/27 If a host name lookup failed very early in a connection, for example, if
4837 the IP address matched host_lookup and the reverse lookup yielded a name
4838 that did not have a forward lookup, an error message of the form "no IP
4839 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4840 could be logged. Now it outputs the IP address instead of "NULL".
4842 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4843 allows the sender and the authenticated sender to be set when
4844 submitting a message from within Exim. Since child_open_exim() is
4845 documented for local_scan(), the new function should be too.
4847 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4848 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4849 results in an empty string is now treated as unset.
4851 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4853 PH/31 Added sender_verify_fail logging option.
4855 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4856 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4857 tidied the source and removed it altogether.
4859 PH/33 When a queue run was abandoned because the load average was too high, a
4860 log line was always written; now it is written only if the queue_run log
4861 selector is set. In addition, the log line for abandonment now contains
4862 information about the queue run such as the pid. This is always present
4863 in "start" and "stop" lines but was omitted from the "abandon" line.
4865 PH/34 Omit spaces between a header name and the colon in the error message that
4866 is given when verify = headers_syntax fails (if there are lots of them,
4867 the message gets confusing).
4869 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4870 names, as there are now some PTR records that contain slashes. This check
4871 is only to protect against broken name servers that fall over on strange
4872 characters, so the fact that it applies to all lookups doesn't matter.
4874 PH/36 Now that the new test suite is complete, we can remove some of the
4875 special code in Exim that was needed for the old test suite. For example,
4876 sorting DNS records because real resolvers return them in an arbitrary
4877 order. The new test suite's fake resolver always returns records in the
4880 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4881 bounces) except when queue_only is set, to avoid logging races between
4882 the different processes.
4884 PH/38 Panic-die if .include specifies a non-absolute path.
4886 PH/39 A tweak to the "H" retry rule from its user.
4888 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4889 a label. They prevented compilation on older perls.
4891 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4892 a warning to be raised on newish perls.
4894 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4895 on queue. Changes to match documented behaviour of showing count of
4896 messages matching specified criteria.
4898 PH/40 Changed the default ident timeout from 30s to 5s.
4900 PH/41 Added support for the use of login_cap features, on those BSD systems
4901 that have them, for controlling the resources used by pipe deliveries.
4903 PH/42 The content-scanning code uses fopen() to create files in which to put
4904 message data. Previously it was not paying any attention to the mode of
4905 the files. Exim runs with umask(0) because the rest of the code creates
4906 files with open(), and sets the required mode explicitly. Thus, these
4907 files were ending up world-writeable. This was not a big issue, because,
4908 being within the spool directory, they were not world-accessible. I have
4909 created a function called modefopen, which takes an additional mode
4910 argument. It sets umask(777), creates the file, chmods it to the required
4911 mode, then resets the umask. All the relevant calls to fopen() in the
4912 content scanning code have been changed to use this function.
4914 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4915 to 24 hours. This avoids potential overflow problems when processing G
4916 and H retry rules. I suspect nobody ever tinkers with this value.
4918 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4920 PH/45 When the plaintext authenticator is running as a client, the server's
4921 challenges are checked to ensure they are valid base64 strings. By
4922 default, the authentication attempt is cancelled if an invalid string is
4923 received. Setting client_ignore_invalid_base64 true ignores these errors.
4924 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4925 they are received. Thus, the responses can be made to depend on the
4926 challenges. If an invalid string is ignored, an empty string is placed in
4929 PH/46 Messages that are created by the autoreply transport now contains a
4930 References: header, in accordance with RFCs 2822 and 3834.
4932 PH/47 Added authenticated_sender_force to the smtp transport.
4934 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4936 PH/49 Installed latest patch from the Sieve maintainer.
4938 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4939 was also set, the appendfile transport was unnecessarily scanning a
4940 directory of message files (e.g. for maildir delivery) to find the count
4941 of files (along with the size), even though it did not need this
4942 information. It now does the scan only if it needs to find either the
4943 size of the count of files.
4945 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4947 PH/52 Two bugs concerned with error handling when the smtp transport is
4950 (i) Exim was not creating retry information for temporary errors given
4951 for individual recipients after the DATA command when the smtp transport
4952 was used in LMTP mode. This meant that they could be retried too
4953 frequently, and not timed out correctly.
4955 (ii) Exim was setting the flag that allows error details to be returned
4956 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4957 recipients that were returned after the DATA command.
4959 PH/53 This is related to PH/52, but is more general: for any failing address,
4960 when detailed error information was permitted to be returned to the
4961 sender, but the error was temporary, then after the final timeout, only
4962 "retry timeout exceeded" was returned. Now it returns the full error as
4963 well as "retry timeout exceeded".
4965 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4966 do this, and (what is worse) MTAs that accept it.
4968 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4969 will now be deprecated.
4971 PH/56 New os.c-cygwin from the Cygwin maintainer.
4973 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4974 formats (previously only available in exim formats via -bpr, -bpru,
4975 and -bpra. Now also available in native and exiqgrep formats)
4977 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4978 with very large, slow to parse queues
4980 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4982 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4984 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4985 responses to authentication challenges, though it was showing the
4986 challenges; (ii) I've removed the CR characters from the debug output for
4989 PH/58 Allow for the insertion of a newline as well as a space when a string
4990 is turned into more than one encoded-word during RFC 2047 encoding. The
4991 Sieve code now uses this.
4993 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4994 data_4xx, lost_connection, tls_required.
4996 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4997 message was being sent as an SMTP response.
4999 PH/61 Add -l and -k options to exicyclog.
5001 PH/62 When verifying, if an address was redirected to one new address, so that
5002 verification continued, and the new address failed or deferred after
5003 having set something in $address_data, the value of $address_data was not
5004 passed back to the ACL. This was different to the case when no
5005 redirection occurred. The value is now passed back in both cases.
5007 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5008 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5009 use login_cap.h, so on its own it isn't the distinguishing feature. The
5010 new name refers directly to the setclassresources() function.
5012 PH/65 Added configuration files for NetBSD3.
5014 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5016 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5017 is preferred over IPv4.
5019 PH/68 The bounce_return_message and bounce_return_body options were not being
5020 honoured for bounces generated during the reception of non-SMTP messages.
5021 In particular, this applied to messages rejected by the ACL. This bug has
5022 been fixed. However, if bounce_return_message is true and bounce_return_
5023 body is false, the headers that are returned for a non-SMTP message
5024 include only those that have been read before the error was detected.
5025 (In the case of an ACL rejection, they have all been read.)
5027 PH/69 The HTML version of the specification is now built in a directory called
5028 spec_html instead of spec.html, because the latter looks like a path with
5029 a MIME-type, and this confuses some software.
5031 PH/70 Catch two compiler warnings in sieve.c.
5033 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5034 function verify_get_ident() calls ip_connect() to connect a socket, but
5035 if the "connect()" function timed out, ip_connect() used to close the
5036 socket. However, verify_get_ident() also closes the socket later, and in
5037 between Exim writes to the log, which may get opened at this point. When
5038 the socket was closed in ip_connect(), the log could get the same file
5039 descriptor number as the socket. This naturally causes chaos. The fix is
5040 not to close the socket in ip_connect(); the socket should be closed by
5041 the function that creates it. There was only one place in the code where
5042 this was missing, in the iplookup router, which I don't think anybody now
5043 uses, but I've fixed it anyway.
5045 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5046 well as to direct DNS lookups. Otherwise the handling of names in host
5047 lists is inconsistent and therefore confusing.
5053 PH/01 Two changes to the default runtime configuration:
5055 (1) Move the checks for relay_from_hosts and authenticated clients from
5056 after to before the (commented out) DNS black list checks.
5058 (2) Add control=submission to the relay_from_hosts and authenticated
5059 clients checks, on the grounds that messages accepted by these
5060 statements are most likely to be submissions.
5062 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5064 (1) Generate an error if the third argument for the ${prvs expansion is
5067 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5070 (3) Reset the variables that are obtained from the first argument of
5071 ${prvscheck and used in the second argument before leaving the code,
5072 because their memory is reclaimed, so using them afterwards may do
5075 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5076 one (it's much easier than Tom thought :-).
5078 (5) Because of (4), we can now allow for the use of $prvscheck_result
5079 inside the third argument.
5081 PH/03 For some reason, the default setting of PATH when running a command from
5082 a pipe transport was just "/usr/bin". I have changed it to
5085 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5086 anything to be listed in the output from -bV.
5088 PH/05 When a filter generated an autoreply, the entire To: header line was
5089 quoted in the delivery log line, like this:
5091 => >A.N.Other <ano@some.domain> <original@ddress> ...
5093 This has been changed so that it extracts the operative address. There
5094 may be more than one such address. If so, they are comma-separated, like
5097 => >ano@some.domain,ona@other.domain <original@ddress> ...
5099 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5100 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5101 not being looked up in the rDNS to get a host name, Exim was showing the
5102 IP address twice in Received: lines, even though the IP addresses were
5103 identical. For example:
5105 Received: from [1.2.3.4] (helo=[1.2.3.4])
5107 However, if the real host name was known, it was omitting the HELO data
5108 if it matched the actual IP address. This has been tidied up so that it
5109 doesn't show the same IP address twice.
5111 PH/07 When both +timestamp and +memory debugging was on, the value given by
5112 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5113 called by the debug printing, thereby overwriting the timestamp buffer.
5114 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5116 PH/08 When the original message was included in an autoreply transport, it
5117 always said "this is a copy of the message, including all the headers",
5118 even if body_only or headers_only was set. It now gives an appropriate
5121 PH/09 Applied a patch from the Sieve maintainer which:
5123 o fixes some comments
5124 o adds the (disabled) notify extension core
5125 o adds some debug output for the result of if/elsif tests
5126 o points to the current vacation draft in the documentation
5127 and documents the missing references header update
5131 o fixes a bug in processing the envelope test (when testing
5132 multiple envelope elements, the last element determined the
5135 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5136 Electronic Mail") by including:
5138 Auto-submitted: auto-generated
5140 in the messages that it generates (bounce messages and others, such as
5141 warnings). In the case of bounce messages for non-SMTP messages, there was
5142 also a typo: it was using "Auto_submitted" (underscore instead of
5143 hyphen). Since every message generated by Exim is necessarily in response
5144 to another message, thes have all been changed to:
5146 Auto-Submitted: auto-replied
5148 in accordance with these statements in the RFC:
5150 The auto-replied keyword:
5152 - SHOULD be used on messages sent in direct response to another
5153 message by an automatic process,
5155 - MUST NOT be used on manually-generated messages,
5157 - MAY be used on Delivery Status Notifications (DSNs) and Message
5158 Disposition Notifications (MDNs),
5160 - MUST NOT be used on messages generated by automatic or periodic
5161 processes, except for messages which are automatic responses to
5164 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5165 to the default Received: header definition.
5167 PH/12 Added log selector acl_warn_skipped (default on).
5169 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5170 variables because (a) they are in the wrong storage pool and (b) even if
5171 they were copied, it wouldn't work properly because of the caching.
5173 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5174 checking when decoding. Apparently there are clients that generate
5175 overlong encoded strings. Why am I not surprised?
5177 PH/15 If the first argument of "${if match_address" was not empty, but did not
5178 contain an "@" character, Exim crashed. Now it writes a panic log message
5179 and treats the condition as false.
5181 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5183 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5184 extension "envelope-auth". The code is finished and in agreement with
5185 other implementations, but there is no documentation so far and in fact,
5186 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5187 not changing the active code.
5189 Print executed "if" and "elsif" statements when debugging is used. This
5190 helps a great deal to understand what a filter does.
5192 Document more things not specified clearly in RFC3028. I had all this
5193 sorted out, when out of a sudden new issues came to my mind. Oops."
5195 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5198 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5199 canonical form (as documented). However, after a host name lookup from
5200 the IP address, check_host() was doing a simple string comparison with
5201 addresses acquired from the DNS when checking that the found name did
5202 have the original IP as one of its addresses. Since any found IPv6
5203 addresses are likely to be in abbreviated form, the comparison could
5204 fail. Luckily, there already exists a function for doing the comparison
5205 by converting both addresses to binary, so now that is used instead of
5206 the text comparison.
5208 PH/20 There was another similar case to PH/19, when a complete host name was
5209 given in a host list; looking up its IP address could give an abbreviated
5210 form, whereas the current host's name might or might not be abbreviated.
5211 The same fix has been applied.
5217 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5218 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5221 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5222 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5224 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5226 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5227 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5228 Also, for "Auto-Submitted", treat anything other than "no" as
5229 non-personal, in accordance with RFC 3834. (Previously it treated
5230 anything starting "auto-" as non-personal.)
5232 TF/01 The control=submission/name=... option had a problem with syntax
5233 errors if the name included a slash character. The /name= option
5234 now slurps the rest of the string, so it can include any characters
5235 but it must come last in the list of options (after /sender_retain
5238 PH/05 Some modifications to the interface to the fake nameserver for the new
5246 TK/01 Added the "success_on_redirect" address verification option. See
5247 NewStuff for rationale and an example.
5249 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5251 PH/02 Patch to exigrep to allow it to work on syslog lines.
5253 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5254 fread() to skip over the body file's header line, because in Cygwin the
5255 header line is locked and is inaccessible.
5257 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5258 co-exist for some time) to make it clear that it is the Exim ID that is
5259 referenced, not the Message-ID: header line.
5261 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5262 string_format() function, because snprintf() does not exist on all
5265 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5266 ${stat: expansion item.
5268 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5269 protocol synchronization error", to keep the pedants happy.
5271 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5272 well as for IRIX systems, when gcc is being used. See the host.c source
5275 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5277 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5280 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5281 transport and to the smtp transport in LMTP mode.
5283 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5285 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5286 run a filter in a subprocess. This could lead to confusion in subsequent
5287 lookups in the parent process. There should also be a search_tidyup() at
5288 the end of the subprocess.
5290 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5291 only if the host matched helo_try_verify_hosts, which caused the
5292 verification to occur when the EHLO/HELO command was issued. The ACL just
5293 tested the remembered result. Now, if a previous verification attempt has
5294 not happened, "verify = helo" does it there and then.
5296 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5298 TK/03 Fix log output including CR from clamd.
5300 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5301 exist provoked a memory error which could cause a segfault.
5303 PH/15 Installed PCRE 6.2
5305 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5307 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5308 of the problem. Specifically, suggested +O2 rather than +O1 for the
5311 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5313 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5314 record, the port was not correctly logged when the outgoing_port log
5315 selector was set (it logged the transort's default port).
5317 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5318 fallback_hosts, and "hosts" in the smtp transport.
5320 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5321 host errors such as "Connection refused".
5323 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5324 authentication with radiusclient 0.4.9:
5326 - Error returned from rc_read_config was caught wrongly
5327 - Username/password not passed on to radius server due to wrong length.
5329 The presumption is that some radiusclient API changes for 4.51/PH/17
5330 were not taken care of correctly. The code is still untested by me (my
5331 Linux distribution still has 0.3.2 of radiusclient), but it was
5332 contributed by a Radius user.
5334 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5335 expanding the "port" option of the smtp transport.
5337 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5338 while reading a MIME header. Thanks to Tom Hughes for a patch.
5340 PH/24 Include config.h inside local_scan.h so that configuration settings are
5343 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5344 that in an ACL for RCPT (for example), you can examine exactly what was
5347 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5348 commands, but it was not correctly comparing the address with the actual
5349 client host address. Thus, it would show the EHLO address in Received:
5350 header lines when this was not necessary.
5352 PH/27 Added the % operator to ${eval:}.
5354 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5355 it should be ignoring failures (because with -C, for example, it has lost
5356 privilege). It wasn't ignoring creation failures other than "already
5359 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5362 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5363 because an input line was too long, either on its own, or by virtue of
5364 too many continuations, the temporary file was not being removed, and the
5365 return code was incorrect.
5367 PH/31 Missing "BOOL" in function definition in filtertest.c.
5369 PH/32 Applied Sieve patches from the maintainer.
5371 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5373 PH/33 Added "verify = not_blind".
5375 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5376 Local/Makefile (with some defaults set). These are used in built scripts
5377 such as exicyclog, but they have never been used in the exim_install
5378 script (though there are many overriding facilities there). I have
5379 arranged that the exim_install script now takes note of these two
5382 PH/35 Installed configuration files for Dragonfly.
5384 PH/36 When a locally submitted message by a trusted user did not contain a
5385 From: header, and the sender address was obtained from -f or from an SMTP
5386 MAIL command, and the trusted user did not use -F to supply a sender
5387 name, $originator_name was incorrectly used when constructing a From:
5388 header. Furthermore, $originator_name was used for submission mode
5389 messages from external hosts without From: headers in a similar way,
5390 which is clearly wrong.
5392 PH/37 Added control=suppress_local_fixups.
5394 PH/38 When log_selector = +received_sender was set, and the addition of the
5395 sender made the log line's construction buffer exactly full, or one byte
5396 less than full, an overflow happened when the terminating "\n" was
5399 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5400 when the result of a list match is failure because a DNS lookup failed.
5402 PH/40 RM_COMMAND is now used in the building process.
5404 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5405 the "build-* directories that it finds.
5407 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5408 address in a domain literal was a prefix of an interface address.
5410 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5411 when verifying a sender address, unless rewrite_headers is false.
5413 PH/44 Wrote a long comment about why errors_to addresses are verified as
5414 recipients, not senders.
5416 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5417 the ratelimit ACL was added.
5419 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5421 PH/46 Added extra information about PostgreSQL errors to the error string.
5423 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5424 avoiding the need to install special zones in a real server. This is
5425 backwards compatible; if it can't find the fake resolver, it drops back.
5426 Thus, both old and new test suites can be run.
5428 TF/02 Added util/ratelimit.pl
5430 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5431 clock is set back in time.
5433 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5434 Candler <B.Candler@pobox.com>.
5436 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5437 OK for addresses that are the result of redirections.
5439 PH/48 A number of further additions for the benefit of the new test suite,
5440 including a fake gethostbyname() that interfaces to the fake DNS resolver
5443 TF/06 The fix for widen_domains has also been applied to qualify_single and
5444 search_parents which are the other dnslookup options that can cause
5447 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5450 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5452 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5453 service" if it encountered an MX record with an empty target hostname.
5454 The message is now "an MX or SRV record indicated no SMTP service".
5456 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5457 if the DNS of the sending site is misconfigured. This is quite a
5458 common situation. This change restores the behaviour of treating a
5459 helo verification defer as a failure.
5461 PH/51 If self=fail was set on a router, the bounce message did not include the
5462 actual error message.
5468 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5470 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5471 command itself ended in error, the underlying message about the transport
5472 filter timeout was being overwritten with the pipe command error. Now the
5473 underlying error message should be appended to the second error message.
5475 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5477 PH/02 Reduce the amount of output that "make" produces by default. Full output
5478 can still be requested.
5480 PH/03 The warning log line about a condition test deferring for a "warn" verb
5481 was being output only once per connection, rather than after each
5482 occurrence (because it was using the same function as for successful
5483 "warn" verbs). This seems wrong, so I have changed it.
5485 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5486 it should not have, which might have caused a crash in the right
5487 circumstances, but probably never did.
5489 PH/04 Installed a modified version of Tony Finch's patch to make submission
5490 mode fix the return path as well as the Sender: header line, and to
5491 add a /name= option so that you can make the user's friendly name appear
5494 TF/03 Added the control = fakedefer ACL modifier.
5496 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5497 Mark Lowes for thorough testing.
5499 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5501 TK/03 Merged latest SRS patch from Miles Wilton.
5503 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5504 with the definition in sysexits.h (which is #included earlier).
5505 Fortunately, Exim does not actually use EX_OK. The code used to try to
5506 preserve the sysexits.h value, by assuming that macro definitions were
5507 scanned for macro replacements. I have been disabused of this notion,
5508 so now the code just undefines EX_OK before #including unistd.h.
5510 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5511 in the smtp transport. When a block could not be written in a single
5512 write() function, the timeout was being re-applied to each part-write.
5513 This seems wrong - if the receiver was accepting one byte at a time it
5514 would take for ever. The timeout is now adjusted when this happens. It
5515 doesn't have to be particularly precise.
5517 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5518 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5520 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5521 without a domain if the check to <postmaster@domain> fails.
5523 SC/01 Eximstats: added -xls and the ability to specify output files
5524 (patch written by Frank Heydlauf).
5526 SC/02 Eximstats: use FileHandles for outputting results.
5528 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5530 SC/04 Eximstats: fixed display of large numbers with -nvr option
5532 SC/05 Eximstats: fixed merging of reports with empty tables.
5534 SC/06 Eximstats: added the -include_original_destination flag
5536 SC/07 Eximstats: removed tabs and trailing whitespace.
5538 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5540 TK/06 MBOX spool code: Add real "From " MBOX separator line
5541 so the .eml file is really in mbox format (even though
5542 most programs do not really care). Patch from Alex Miller.
5544 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5545 The latter is generated from $received_to and is only set if the
5546 message has one envelope recipient. SA can use these headers,
5547 obviously out-of-the-box. Patch from Alex Miller.
5549 PH/08 The ${def test on a variable was returning false if the variable's
5550 value was "0", contrary to what the specification has always said!
5551 The result should be true unless the variable is empty.
5553 PH/09 The syntax error of a character other than { following "${if
5554 def:variable_name" (after optional whitespace) was not being diagnosed.
5555 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5556 accidental colon was present, for example, could give incorrect results.
5558 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5559 result is used (not including appendfile, where other changes are about
5562 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5563 This involved changing a lot of size variables from int to off_t. It
5564 should work with maildirs and everything.
5566 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5567 spamd dying while we are connected to it.
5569 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5572 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5573 function for BDB 4.3.
5575 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5577 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5578 the sources for PCRE have been re-arranged and more files are now
5581 PH/15 The code I had for printing potentially long long variables in PH/11
5582 above was not the best (it lost precision). The length of off_t variables
5583 is now inspected at build time, and an appropriate printing format (%ld
5584 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5585 to be "long long int" or "long int". This is needed for the internal
5586 formatting function string_vformat().
5588 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5589 the configuration file to be ":syslog", then the script "guesses" where
5590 the logs files are, rather than using the compiled in default. In our
5591 case the guess is not the same as the compiled default, so the script
5592 suddenly stopped working when I started to use syslog. The patch checks
5593 to see if log_file_path is "". If so, it attempts to read it from exim
5594 with no configuration file to get the compiled in version, before it
5595 falls back to the previous guessing code."
5597 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5598 implementing BATV in an Exim configuration. See NewStuff for the gory
5601 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5602 Makefile that are specific to HP-UX.
5604 PH/18 If the "use_postmaster" option was set for a recipient callout together
5605 with the "random" option, the postmaster address was used as the MAIL
5606 FROM address for the random test, but not for the subsequent recipient
5607 test. It is now used for both.
5609 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5610 patch removes a few documentation additions to RFC 3028, because the
5611 latest draft now contains them. It adds the new en;ascii-case comparator
5612 and a new error check for 8bit text in MIME parts. Comparator and
5613 require names are now matched exactly. I enabled the subaddress
5614 extension, but it is not well tested yet (read: it works for me)."
5616 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5617 rework some of the code of TK/09 above to avoid the hardwired use of
5618 "%lld" and "long long". Replaced the call to snprintf() with a call to
5621 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5622 records point to non-existent hosts", "retry timeout exceeded", and
5623 "retry time not reached for any host after a long failure period".
5625 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5626 experimental DomainKeys support:
5628 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5629 (2) On an error such as an illegally used "control", the wrong name for
5630 the control was given.
5632 These problems did NOT occur unless DomainKeys support was compiled.
5634 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5636 PH/24 Added ${if match_ip condition.
5638 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5639 databases so that it will be absolutely obvious if a crash occurs in the
5640 DB library. This is a regular occurrence (often caused by mis-matched
5643 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5644 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5645 picked up on a user's system that detects such things. There doesn't seem
5646 to be a gcc warning option for this - only an attribute that has to be
5647 put on the function's prototype. It seems that in Fedora Core 4 they have
5648 set this on a number of new functions. No doubt there will be more in due
5651 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5652 specify a transport. However, if an address that was verified by such a
5653 router was the subject of a callout, Exim crashed because it tried to
5654 read the rcpt_include_affixes from the non-existent transport. Now it
5655 just assumes that the setting of that option is false. This bug was
5656 introduced by 4.51/PH/31.
5658 PH/28 Changed -d+all to exclude +memory, because that information is very
5659 rarely of interest, but it makes the output a lot bigger. People tend to
5660 do -d+all out of habit.
5662 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5663 code in os-type was giving problems when libc.so lives in lib64, like on
5666 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5667 aren't the modern standard, and it seems that some systems' include files
5668 don't always have them. Exim was already checking for some of the newer
5669 ones like T_AAAA, and defining it itself. I've added checks for all the
5670 record types that Exim uses.
5672 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5673 not automatically generating a new one, as it is supposed to. This
5674 prevented TLS from working. If the file did exist, but contained invalid
5675 data, a new version was generated, as expected. It was only the case of a
5676 non-existent file that was broken.
5678 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5679 with a change in libdomainkeys > 0.64.
5681 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5682 from DNS. If the selector record carries the flag, it now has
5683 precedence over the domain-wide flag.
5685 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5687 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5688 the use of a port name that isn't defined in /etc/services) occurred, the
5689 message was deferred as in a normal delivery, and thus remained on the
5690 spool, instead of being failed because of the mua_wrapper setting. This
5691 is now fixed, and I tidied up some of the mua_wrapper messages at the
5694 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5695 the messages in a hash of arrays rather than using individual hashes.
5696 This is a bit cleaner and results in dramatic memory savings, albeit
5697 at a slight CPU cost.
5699 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5700 as requested by Marc Sherman.
5702 SC/10 Eximstats: added histograms for user specified patterns as requested
5705 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5707 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5708 fopen() in the content-scanning modules that did not already have it.
5714 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5715 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5717 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5719 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5721 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5722 to test Sieve filters that use "vacation".
5724 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5725 that changes the way the GnuTLS parameters are stored in the cache file.
5726 The new format can be generated externally. For backward compatibility,
5727 if the data in the cache doesn't make sense, Exim assumes it has read an
5728 old-format file, and it generates new data and writes a new file. This
5729 means that you can't go back to an older release without removing the
5732 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5733 work if there are any delivery delays because "one_time" forces the
5734 parent to be marked "delivered", so its unseen clone is never tried
5735 again. For this reason, Exim now forbids the simultaneous setting of
5738 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5739 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5740 entry if you want to know the details. The fix, however, neglected to
5741 consider the case when local delivery batching is involved. The test for
5742 "previously delivered" was not happening when checking to see if an
5743 address could be batched with a previous (undelivered) one; under
5744 certain circumstances this could lead to multiple deliveries to the same
5747 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5748 in its include files, and this causes problems building Exim.
5750 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5751 header_syntax) but Exim was just ignoring anything given after a slash.
5752 In particular, this caused confusion with an attempt to use "verify =
5753 reverse_host_lookup/defer_ok". An error is now given when options are
5754 supplied for verify items that do not have them. (Maybe reverse_host_
5755 lookup should have a defer_ok option, but that's a different point.)
5757 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5758 defined by RFC 821) to 2048, because there were problems with some AUTH
5759 commands, and RFC 1869 says the size should be increased for extended
5760 SMTP commands that take arguments.
5762 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5765 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5766 "unknown" error; now it says that the functionality isn't in the binary.
5768 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5769 an address' error message when a string expansion fails (syntax or
5770 whatever). Otherwise the password may appear in the log. Following change
5771 PH/42 below, there is no longer a chance of it appearing in a bounce
5774 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5776 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5777 of its addresses was ever tried. (Bugzilla bug #2.)
5779 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5780 the result incorrectly in the debug output. (It correctly added a newline
5781 to what was transported.)
5783 TF/01 Added $received_time.
5785 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5786 commented out examples of how to interface to a virus scanner and to
5787 SpamAssassin. Also added commented examples of av_scanner and
5788 spamd_address settings.
5790 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5791 and controls are allowed in which ACLs. There were a couple of minor
5792 errors. Some of the entries in the conditions table (which is a table of
5793 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5794 negation of where the condition IS allowed.
5796 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5798 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5799 header file does not have a version number, so I've had to invent a new
5800 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5801 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5802 radiusclient), but it was contributed by a Radius user.
5804 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5805 files or directories, for OpenSSL.
5807 PH/19 When an Exim process that is running as root has to create an Exim log
5808 file, it does so in a subprocess that runs as exim:exim so as to get the
5809 ownership right at creation (otherwise, other Exim processes might see
5810 the file with the wrong ownership). There was no test for failure of this
5811 fork() call, which would lead to the process getting stuck as it waited
5812 for a non-existent subprocess. Forks do occasionally fail when resources
5813 run out. I reviewed all the other calls to fork(); they all seem to check
5816 PH/20 When checking for unexpected SMTP input at connect time (before writing
5817 the banner), Exim was not dealing correctly with a non-positive return
5818 from the read() function. If the client had disconnected by this time,
5819 the result was a log entry for a synchronization error with an empty
5820 string after "input=" when read() returned zero. If read() returned -1
5821 (an event I could not check), uninitialized data bytes were printed.
5822 There were reports of junk text (parts of files, etc) appearing after
5825 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5827 PH/22 Added support for macro redefinition, and (re)definition in between
5828 driver and ACL definitions.
5830 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5831 forgetting to use the resulting value; it was using the unexpanded value.
5833 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5834 hadn't been configured. The fix is from Juergen Kreileder, who
5835 understands it better than I do:
5837 "Here's what I see happening with three configured cyrus_sasl
5838 authenticators configured (plain, login, cram-md5):
5840 On startup auth_cyrus_sasl_init() gets called for each of these.
5841 This means three calls to sasl_listmech() without a specified mech_list.
5842 => SASL tests which mechs of all available mechs actually work
5843 => three warnings about OTP not working
5844 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5846 With the patch, sasl_listmech() also gets called three times. But now
5847 SASL's mech_list option is set to the server_mech specified in the the
5848 authenticator. Or in other words, the answer from sasl_listmech()
5849 gets limited to just the mech you're testing for (which is different
5851 => the return list contains just 'plain' or 'login', 'cram-md5' or
5852 nothing depending on the value of ob->server_mech.
5854 I've just tested the patch: Authentication still works fine,
5855 unavailable mechs specified in the exim configuration are still
5856 caught, and the auth.log warnings about OTP are gone."
5858 PH/25 When debugging is enabled, the contents of the command line are added
5859 to the debugging output, even when log_selector=+arguments is not
5862 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5863 answer is "GNU", and only if the return is "GNU/something" is the answer
5866 PH/27 $acl_verify_message is now set immediately after the failure of a
5867 verification in an ACL, and so is available in subsequent modifiers. In
5868 particular, the message can be preserved by coding like this:
5870 warn !verify = sender
5871 set acl_m0 = $acl_verify_message
5873 Previously, $acl_verify_message was set only while expanding "message"
5874 and "log_message" when a very denied access.
5876 PH/28 Modified OS/os.c-Linux with
5878 -#ifndef OS_LOAD_AVERAGE
5879 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5881 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5882 nomenclature these days.)
5884 PH/29 Installed patch from the Sieve maintainer that adds the options
5885 sieve_useraddress and sieve_subaddress to the redirect router.
5887 PH/30 In these circumstances:
5888 . Two addresses routed to the same list of hosts;
5889 . First host does not offer TLS;
5890 . First host accepts first address;
5891 . First host gives temporary error to second address;
5892 . Second host offers TLS and a TLS session is established;
5893 . Second host accepts second address.
5894 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5895 and peerdn, if requested) that were in fact used only for the second
5898 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5899 attention to any local part prefix or suffix that was matched by the
5900 router that accepted the address. It now behaves in the same way as it
5901 does for delivery: the affixes are removed from the local part unless
5902 rcpt_include_affixes is set on the transport.
5904 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5905 timeout during the DATA phase of an incoming message.
5907 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5908 applied a patch sanctioned by the Sieve maintainer.
5910 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5911 the uid or gid is negative. A case of a negative gid caused this to be
5912 noticed. The fix allows for either to be negative.
5914 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5915 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5918 PH/36 The change PH/12 above was broken. Fixed it.
5920 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5921 the grounds that routing the same address twice would always produce the
5922 same answer. This might have been true once, but it is certainly no
5923 longer true now. Routing a child address may depend on the previous
5924 routing that produced that child. Some complicated redirection strategies
5925 went wrong when messages had multiple recipients, and made Exim's
5926 behaviour dependent on the order in which the addresses were given.
5928 I have moved the duplicate checking until after the routing is complete.
5929 Exim scans the addresses that are assigned to local and remote
5930 transports, and removes any duplicates. This means that more work will be
5931 done, as duplicates will always all be routed, but duplicates are
5932 presumably rare, so I don't expect this is of any significance.
5934 For deliveries to pipes, files, and autoreplies, the duplicate checking
5935 still happens during the routing process, since they are not going to be
5938 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5939 It corrects a timeout issue with spamd. This is Ian's comment: "The
5940 background is that sometimes spamd either never reads data from a
5941 connection it has accepted, or it never writes response data. The exiscan
5942 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5943 blindly assumes that writes won't block so it may never time out."
5945 PH/39 Allow G after quota size as well as K and M.
5947 PH/40 The value set for $authenticated_id in an authenticator may not contain
5948 binary zeroes or newlines because the value is written to log lines and
5949 to spool files. There was no check on this. Now the value is run through
5950 the string_printing() function so that such characters are converted to
5951 printable escape sequences.
5953 PH/41 $message_linecount is a new variable that contains the total number of
5954 lines in the message. Compare $body_linecount, which is the count for the
5957 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5958 bounce and delay warning messages, except in certain special cases, which
5961 (a) An SMTP error message from a remote host;
5962 (b) A message specified in a :fail: redirection;
5963 (c) A message specified in a "fail" command in a system filter;
5964 (d) A message specified in a FAIL return from the queryprogram router;
5965 (e) A message specified by the cannot_route_message router option.
5967 In these cases only, Exim does include the error details in bounce and
5968 warning messages. There are also a few cases where bland messages such
5969 as "unrouteable address" or "local delivery error" are given.
5971 PH/43 $value is now also set for the "else" part of a ${run expansion.
5973 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5974 being worked on, but at least Exim now implements the latest version to
5977 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5978 process to complete was treated as a delivery failure, a timeout while
5979 writing the message to the pipe was logged, but erroneously treated as a
5980 successful delivery. Such timeouts include transport filter timeouts. For
5981 consistency with the overall process timeout, these timeouts are now
5982 treated as errors, giving rise to delivery failures by default. However,
5983 there is now a new Boolean option for the pipe transport called
5984 timeout_defer, which, if set TRUE, converts the failures into defers for
5985 both kinds of timeout. A transport filter timeout is now identified in
5988 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5989 systems where "make" and "gmake" are different, calling "gmake" at top
5990 level broke things. I've arranged for the value of $(MAKE) to be passed
5991 from the Makefile to this script so that it can call the same version of
5995 A note about Exim versions 4.44 and 4.50
5996 ----------------------------------------
5998 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5999 changes of various kinds. As a consequence, a big documentation update was
6000 needed. This delayed the release for rather longer than seemed good, especially
6001 in the light of a couple of (minor) security issues. Therefore, the changes
6002 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6003 release. So 4.44 and 4.50 are in effect two different branches that both start
6006 I have left the 4.50 change log unchanged; it contains all the changes since
6007 4.43. The change log for 4.44 is below; many of its items are identical to
6008 those for 4.50. This seems to be the most sensible way to preserve the
6009 historical information.
6015 1. Minor wording change to the doc/README.SIEVE file.
6017 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6018 computation of the current number of files was incorrect.
6020 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6021 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6024 4. Give more explanation in the error message when the command for a transport
6025 filter fails to execute.
6027 5. There are several places where Exim runs a non-Exim command in a
6028 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6029 was being done only for the command run by the queryprogram router. It is
6030 now done for all such subprocesses. The other cases are: ${run, transport
6031 filters, and the commands run by the lmtp and pipe transports.
6033 6. Added CONFIGURE_GROUP build-time option.
6035 7. Some older OS have a limit of 256 on the maximum number of file
6036 descriptors. Exim was using setrlimit() to set 1000 as a large value
6037 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6038 systems. I've change it so that if it can't get 1000, it tries for 256.
6040 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6041 was an oversight, and furthermore, ever since the addition of extra
6042 controls (e.g. 4.43/32), the checks on when to allow different forms of
6043 "control" were broken. There should now be diagnostics for all cases when a
6044 control that does not make sense is encountered.
6046 9. Added the /retain_sender option to "control=submission".
6048 10. $recipients is now available in the predata ACL (oversight).
6050 11. Tidy the search cache before the fork to do a delivery from a message
6051 received from the command line. Otherwise the child will trigger a lookup
6052 failure and thereby defer the delivery if it tries to use (for example) a
6053 cached ldap connection that the parent has called unbind on.
6055 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6056 of $address_data from the recipient verification was clobbered by the
6057 sender verification.
6059 13. The value of address_data from a sender verification is now available in
6060 $sender_address_data in subsequent conditions in the ACL statement.
6062 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6064 15. Added a new option "connect=<time>" to callout options, to set a different
6067 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6068 was its contents. (It was OK if the option was not defined at all.)
6070 17. A "Completed" log line is now written for messages that are removed from
6071 the spool by the -Mrm option.
6073 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6074 information about exactly what failed.
6076 19. Added -dd to debug only the daemon process.
6078 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6079 handles timeouts, both on the server side and network timeouts. Renamed the
6080 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6082 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6083 It is now set to "smtps".
6085 22. $host_address is now set to the target address during the checking of
6086 ignore_target_hosts.
6088 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6089 being passed; this would have caused $sender_host_name to have been used if
6090 matching the list had actually called for a host name (not very likely,
6091 since this list is usually IP addresses). A host name is now passed as
6094 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6095 code that specifies a non-restarting handler (typically sigaction() in
6096 modern systems) in an attempt to fix a rare and obscure crash bug.
6098 25. Narrowed the window for a race in the daemon that could cause it to ignore
6099 SIGCHLD signals. This is not a major problem, because they are used only to
6100 wake it up if nothing else does.
6102 26. A malformed maildirsize file could cause Exim to calculate negative values
6103 for the mailbox size or file count. Odd effects could occur as a result.
6104 The maildirsize information is now recalculated if the size or filecount
6107 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6108 support for a long time. Removed HAVE_SYS_VFS_H.
6110 28. Installed the latest version of exipick from John Jetmore.
6112 29. In an address list, if the pattern was not a regular expression, an empty
6113 subject address (from a bounce message) matched only if the pattern was an
6114 empty string. Non-empty patterns were not even tested. This was the wrong
6115 because it is perfectly reasonable to use an empty address as part of a
6116 database query. An empty address is now tested by patterns that are
6117 lookups. However, all the other forms of pattern expect the subject to
6118 contain a local part and a domain, and therefore, for them, an empty
6119 address still always fails if the pattern is not itself empty.
6121 30. Exim went into a mad DNS loop when attempting to do a callout where the
6122 host was specified on an smtp transport, and looking it up yielded more
6123 than one IP address.
6125 31. Re-factored the code for checking spool and log partition space into a
6126 function that finds that data and another that does the check. The former
6127 is then used to implement four new variables: $spool_space, $log_space,
6128 $spool_inodes, and $log_inodes.
6130 32. The RFC2047 encoding function was originally intended for short strings
6131 such as real names; it was not keeping to the 75-character limit for
6132 encoded words that the RFC imposes. It now respects the limit, and
6133 generates multiple encoded words if necessary. To be on the safe side, I
6134 have increased the buffer size for the ${rfc2047: expansion operator from
6137 33. It is now permitted to omit both strings after an "if" condition; if the
6138 condition is true, the result is "true". As before, when the second string
6139 is omitted, a false condition yields an empty string. This makes it less
6140 cumbersome to write custom ACL and router conditions.
6142 34. Failure to deliver a bounce message always caused it to be frozen, even if
6143 there was an errors_to setting on the router. The errors_to setting is now
6146 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6147 canonical form (fully expanded) before being placed in
6148 $sender_host_address.
6150 36. The table in the code that translates DNS record types into text (T_A to
6151 "A" for instance) was missing entries for NS and CNAME. It is just possible
6152 that this could have caused confusion if both these types were looked up
6153 for the same domain, because the text type is used as part of Exim's
6154 per-process caching. But the chance of anyone hitting this buglet seems
6157 37. The dnsdb lookup has been extended in a number of ways.
6159 (1) There is a new type, "zns", which walks up the domain tree until it
6160 finds some nameserver records. It should be used with care.
6162 (2) There is a new type, "mxh", which is like "mx" except that it returns
6163 just the host names, not the priorities.
6165 (3) It is now possible to give a list of domains (or IP addresses) to be
6166 looked up. The behaviour when one of the lookups defers can be
6167 controlled by a keyword.
6169 (4) It is now possible to specify the separator character for use when
6170 multiple records are returned.
6172 38. The dnslists ACL condition has been extended: it is now possible to supply
6173 a list of IP addresses and/or domains to be looked up in a particular DNS
6176 39. Added log_selector=+queue_time_overall.
6178 40. When running the queue in the test harness, wait just a tad after forking a
6179 delivery process, to get repeatability of debugging output.
6181 41. Include certificate and key file names in error message when GnuTLS fails
6182 to set them up, because the GnuTLS error message doesn't include the name
6183 of the failing file when there is a problem reading it.
6185 42. Allow both -bf and -bF in the same test run.
6187 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6189 44. The "Exiscan patch" is now merged into the mainline Exim source.
6191 45. Sometimes the final signoff response after QUIT could fail to get
6192 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6193 < 0 before doing a fflush(). This bug looks as though it goes back to the
6194 introduction of TLS in release 3.20, but "sometimes" must have been rare
6195 because the tests only now provoked it.
6197 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6198 (this can affect the format of dates).
6200 47. exim_tidydb, when checking for the continued existence of a message for
6201 which it has found a message-specific retry record, was not finding
6202 messages that were in split spool directories. Consequently, it was
6203 deleting retry records that should have stayed in existence.
6205 48. Steve fixed some bugs in eximstats.
6207 49. The SPA authentication driver was not abandoning authentication and moving
6208 on to the next authenticator when an expansion was forced to fail,
6209 contradicting the general specification for all authenticators. Instead it
6210 was generating a temporary error. It now behaves as specified.
6212 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6213 (the order specifies the preference for clients). The order is now AES256,
6214 AES128, 3DES, ARCFOUR128.
6216 51. Small patch to Sieve code - explicitly set From: when generating an
6219 52. Exim crashed if a remote delivery caused a very long error message to be
6220 recorded - for instance if somebody sent an entire SpamAssassin report back
6221 as a large number of 550 error lines. This bug was coincidentally fixed by
6222 increasing the size of one of Exim's internal buffers (big_buffer) that
6223 happened as part of the Exiscan merge. However, to be on the safe side, I
6224 have made the code more robust (and fixed the comments that describe what
6227 53. Now that there can be additional text after "Completed" in log lines (if
6228 the queue_time_overall log selector is set), a one-byte patch to exigrep
6229 was needed to allow it to recognize "Completed" as not the last thing in
6232 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6233 patch that reportedly fixes this has been added. I am not expert enough to
6234 create a test for it. This is what the patch creator wrote:
6236 "I found a little strange behaviour of ldap code when working with
6237 Windows 2003 AD Domain, where users was placed in more than one
6238 Organization Units. When I tried to give exim partial DN, the exit code
6239 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6240 But simultaneously result of request was absolutely normal ldap result,
6241 so I produce this patch..."
6243 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6244 REFERENCE, so I have modified the code to exclude the patch when that macro
6247 55. Some experimental protocols are using DNS PTR records for new purposes. The
6248 keys for these records are domain names, not reversed IP addresses. The
6249 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6250 leaves it alone. Component reversal etc. now happens only for IP addresses.
6253 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6255 57. Double the size of the debug message buffer (to 2048) so that more of very
6256 long debug lines gets shown.
6258 58. The exicyclog utility now does better if the number of log files to keep
6259 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6261 59. Two changes related to the smtp_active_hostname option:
6263 (1) $smtp_active_hostname is now available as a variable.
6264 (2) The default for smtp_banner uses $smtp_active_hostname instead
6265 of $primary_hostname.
6267 60. The host_aton() function is supposed to be passed a string that is known
6268 to be a valid IP address. However, in the case of IPv6 addresses, it was
6269 not checking this. This is a hostage to fortune. Exim now panics and dies
6270 if the condition is not met. A case was found where this could be provoked
6271 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6272 components; fortuitously, this particular loophole had already been fixed
6273 by change 4.50/55 above.
6275 If there are any other similar loopholes, the new check in host_aton()
6276 itself should stop them being exploited. The report I received stated that
6277 data on the command line could provoke the exploit when Exim was running as
6278 exim, but did not say which command line option was involved. All I could
6279 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6280 running as the user.
6283 61. There was a buffer overflow vulnerability in the SPA authentication code
6284 (which came originally from the Samba project). I have added a test to the
6285 spa_base64_to_bits() function which I hope fixes it.
6288 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6289 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6291 63. The daemon start-up calls getloadavg() while still root for those OS that
6292 need the first call to be done as root, but it missed one case: when
6293 deliver_queue_load_max is set with deliver_drop_privilege. This is
6294 necessary for the benefit of the queue runner, because there is no re-exec
6295 when deliver_drop_privilege is set.
6297 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6298 This has been fixed.
6300 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6301 was in use, was not putting the data itself into the right store pool;
6302 consequently, it could be overwritten for a subsequent message in the same
6303 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6306 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6308 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6309 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6310 the function were treating the return as a boolean value, which happened to
6311 work because 0=false and not-0=true, but is not correct code.
6313 68. The host_aton() function was not handling scoped IPv6 addresses (those
6314 with, for example, "%eth0" on the end) correctly.
6316 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6317 negated items (that is, ~something) in unsigned ints. Some compilers
6318 apparently mutter when there is no cast.
6320 70. If an address verification called from an ACL failed, and did not produce a
6321 user-specific message (i.e. there was only a "system" message), nothing was
6322 put in $acl_verify_message. In this situation, it now puts the system
6325 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6326 session; change 4.31/43 added the unwanted input to the log line - except
6327 that it did not do this in the start of session case. It now does.
6329 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6330 This is wrong and can cause the other end to generate a synchronization
6331 error if it is another Exim or anything else that does the synchronization
6332 check. A QUIT command is no longer sent after a timeout.
6334 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6335 during host lookups.
6337 74. The defer_ok option of callout verification was not working if it was used
6338 when verifying addresses in header lines, that is, for this case:
6340 verify = header_sender/callout=defer_ok
6342 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6343 those file descriptors could be used for SMTP connections. If anything
6344 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6345 could be sent to the SMTP client, causing chaos. The daemon now opens
6346 stdin, stdout, and stderr to /dev/null when it puts itself into the
6349 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6350 log by default. The user can override this with suitable Perl magic.
6352 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6353 the log message when discard triggers, was not working for the DATA ACL or
6354 for the non-SMTP ACL.
6356 78. Error message wording change in sieve.c.
6358 79. If smtp_accept_max_per_host was set, the number of connections could be
6359 restricted to fewer than expected, because the daemon was trying to set up
6360 a new connection before checking whether the processes handling previous
6361 connections had finished. The check for completed processes is now done
6362 earlier. On busy systems, this bug wouldn't be noticed because something
6363 else would have woken the daemon, and it would have reaped the completed
6366 80. If a message was submitted locally by a user whose login name contained one
6367 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6368 It caused a spool format error. I have fixed the spool reading code. A
6369 related problem was that the "from" clause in the Received: line became
6370 illegal because of the space(s). It is now covered by ${quote_local_part.
6372 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6375 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6377 83. Patch from Sieve maintainer to fix unterminated string problem in
6378 "vacation" handling.
6380 84. Some minor changes to the Linux configuration files to help with other
6381 OS variants using glibc.
6383 85. One more patch for Sieve to update vacation handling to latest spec.
6386 ----------------------------------------------------
6387 See the note above about the 4.44 and 4.50 releases.
6388 ----------------------------------------------------
6394 1. Change 4.43/35 introduced a bug that caused file counts to be
6395 incorrectly computed when quota_filecount was set in an appendfile
6398 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6399 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6402 3. Give more explanation in the error message when the command for a transport
6403 filter fails to execute.
6405 4. There are several places where Exim runs a non-Exim command in a
6406 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6407 was being done only for the command run by the queryprogram router. It is
6408 now done for all such subprocesses. The other cases are: ${run, transport
6409 filters, and the commands run by the lmtp and pipe transports.
6411 5. Some older OS have a limit of 256 on the maximum number of file
6412 descriptors. Exim was using setrlimit() to set 1000 as a large value
6413 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6414 systems. I've change it so that if it can't get 1000, it tries for 256.
6416 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6417 was an oversight, and furthermore, ever since the addition of extra
6418 controls (e.g. 4.43/32), the checks on when to allow different forms of
6419 "control" were broken. There should now be diagnostics for all cases when a
6420 control that does not make sense is encountered.
6422 7. $recipients is now available in the predata ACL (oversight).
6424 8. Tidy the search cache before the fork to do a delivery from a message
6425 received from the command line. Otherwise the child will trigger a lookup
6426 failure and thereby defer the delivery if it tries to use (for example) a
6427 cached ldap connection that the parent has called unbind on.
6429 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6430 of $address_data from the recipient verification was clobbered by the
6431 sender verification.
6433 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6434 was its contents. (It was OK if the option was not defined at all.)
6436 11. A "Completed" log line is now written for messages that are removed from
6437 the spool by the -Mrm option.
6439 12. $host_address is now set to the target address during the checking of
6440 ignore_target_hosts.
6442 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6443 being passed; this would have caused $sender_host_name to have been used if
6444 matching the list had actually called for a host name (not very likely,
6445 since this list is usually IP addresses). A host name is now passed as
6448 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6449 code that specifies a non-restarting handler (typically sigaction() in
6450 modern systems) in an attempt to fix a rare and obscure crash bug.
6452 15. Narrowed the window for a race in the daemon that could cause it to ignore
6453 SIGCHLD signals. This is not a major problem, because they are used only to
6454 wake it up if nothing else does.
6456 16. A malformed maildirsize file could cause Exim to calculate negative values
6457 for the mailbox size or file count. Odd effects could occur as a result.
6458 The maildirsize information is now recalculated if the size or filecount
6461 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6462 support for a long time. Removed HAVE_SYS_VFS_H.
6464 18. Updated exipick to current release from John Jetmore.
6466 19. Allow an empty sender to be matched against a lookup in an address list.
6467 Previously the only cases considered were a regular expression, or an
6470 20. Exim went into a mad DNS lookup loop when doing a callout where the
6471 host was specified on the transport, if the DNS lookup yielded more than
6474 21. The RFC2047 encoding function was originally intended for short strings
6475 such as real names; it was not keeping to the 75-character limit for
6476 encoded words that the RFC imposes. It now respects the limit, and
6477 generates multiple encoded words if necessary. To be on the safe side, I
6478 have increased the buffer size for the ${rfc2047: expansion operator from
6481 22. Failure to deliver a bounce message always caused it to be frozen, even if
6482 there was an errors_to setting on the router. The errors_to setting is now
6485 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6486 canonical form (fully expanded) before being placed in
6487 $sender_host_address.
6489 24. Updated eximstats to version 1.33
6491 25. Include certificate and key file names in error message when GnuTLS fails
6492 to set them up, because the GnuTLS error message doesn't include the name
6493 of the failing file when there is a problem reading it.
6495 26. Expand error message when OpenSSL has problems setting up cert/key files.
6498 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6499 (this can affect the format of dates).
6501 28. exim_tidydb, when checking for the continued existence of a message for
6502 which it has found a message-specific retry record, was not finding
6503 messages that were in split spool directories. Consequently, it was
6504 deleting retry records that should have stayed in existence.
6506 29. eximstats updated to version 1.35
6507 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6508 1.35 - bugfix such that pie charts by volume are generated correctly
6510 30. The SPA authentication driver was not abandoning authentication and moving
6511 on to the next authenticator when an expansion was forced to fail,
6512 contradicting the general specification for all authenticators. Instead it
6513 was generating a temporary error. It now behaves as specified.
6515 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6516 (the order specifies the preference for clients). The order is now AES256,
6517 AES128, 3DES, ARCFOUR128.
6519 31. Small patch to Sieve code - explicitly set From: when generating an
6522 32. Exim crashed if a remote delivery caused a very long error message to be
6523 recorded - for instance if somebody sent an entire SpamAssassin report back
6524 as a large number of 550 error lines. This bug was coincidentally fixed by
6525 increasing the size of one of Exim's internal buffers (big_buffer) that
6526 happened as part of the Exiscan merge. However, to be on the safe side, I
6527 have made the code more robust (and fixed the comments that describe what
6530 33. Some experimental protocols are using DNS PTR records for new purposes. The
6531 keys for these records are domain names, not reversed IP addresses. The
6532 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6533 leaves it alone. Component reversal etc. now happens only for IP addresses.
6536 34. The host_aton() function is supposed to be passed a string that is known
6537 to be a valid IP address. However, in the case of IPv6 addresses, it was
6538 not checking this. This is a hostage to fortune. Exim now panics and dies
6539 if the condition is not met. A case was found where this could be provoked
6540 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6541 components; fortuitously, this particular loophole had already been fixed
6542 by change 4.50/55 or 4.44/33 above.
6544 If there are any other similar loopholes, the new check in host_aton()
6545 itself should stop them being exploited. The report I received stated that
6546 data on the command line could provoke the exploit when Exim was running as
6547 exim, but did not say which command line option was involved. All I could
6548 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6549 running as the user.
6552 35. There was a buffer overflow vulnerability in the SPA authentication code
6553 (which came originally from the Samba project). I have added a test to the
6554 spa_base64_to_bits() function which I hope fixes it.
6557 36. The daemon start-up calls getloadavg() while still root for those OS that
6558 need the first call to be done as root, but it missed one case: when
6559 deliver_queue_load_max is set with deliver_drop_privilege. This is
6560 necessary for the benefit of the queue runner, because there is no re-exec
6561 when deliver_drop_privilege is set.
6563 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6564 was in use, was not putting the data itself into the right store pool;
6565 consequently, it could be overwritten for a subsequent message in the same
6566 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6569 38. Sometimes the final signoff response after QUIT could fail to get
6570 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6571 < 0 before doing a fflush(). This bug looks as though it goes back to the
6572 introduction of TLS in release 3.20, but "sometimes" must have been rare
6573 because the tests only now provoked it.
6579 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6580 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6581 Now it returns an integer. A number of calls were still expecting a T/F
6582 return. Fortuitously, in all cases, the tests worked in OK situations,
6583 which is the norm. However, things would have gone wrong on any write
6584 failures on the smtp file descriptor. This function is used when sending
6585 messages over SMTP and also when doing verify callouts.
6587 2. When Exim is called to do synchronous delivery of a locally submitted
6588 message (the -odf or -odi options), it no longer closes stderr before doing
6591 3. Implemented the mua_wrapper option.
6593 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6595 5. Implemented the functions header_remove(), header_testname(),
6596 header_add_at_position(), and receive_remove_recipient(), and exported them
6599 6. If an ACL "warn" statement specified the addition of headers, Exim already
6600 inserted X-ACL-Warn: at the start if there was no header name. However, it
6601 was not making this test for the second and subsequent header lines if
6602 there were newlines in the string. This meant that an invalid header could
6603 be inserted if Exim was badly configured.
6605 7. Allow an ACL "warn" statement to add header lines at the start or after all
6606 the Received: headers, as well as at the end.
6608 8. Added the rcpt_4xx retry error code.
6610 9. Added postmaster_mailfrom=xxx to callout verification option.
6612 10. Added mailfrom=xxxx to the callout verification option, for verify=
6615 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6616 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6618 12. Inserted some casts to stop certain compilers warning when using pointer
6619 differences as field lengths or precisions in printf-type calls (mostly
6620 affecting debugging statements).
6622 13. Added optional readline() support for -be (dynamically loaded).
6624 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6625 same clock tick as a message's arrival, so that its received time was the
6626 same as the "first fail" time on the retry record, and that message
6627 remained on the queue past the ultimate address timeout, every queue runner
6628 would try a delivery (because it was past the ultimate address timeout) but
6629 after another failure, the ultimate address timeout, which should have then
6630 bounced the address, did not kick in. This was a "< instead of <=" error;
6631 in most cases the first failure would have been in the next clock tick
6632 after the received time, and all would be well.
6634 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6635 being recognized when the domain list was tested by the match_domain
6636 condition in an expansion string.
6638 16. Added the ${str2b64: operator.
6640 17. Exim was always calling setrlimit() to set a large limit for the number of
6641 processes, without checking whether the existing limit was already
6642 adequate. (It did check for the limit on file descriptors.) Furthermore,
6643 errors from getrlimit() and setrlimit() were being ignored. Now they are
6644 logged to the main and panic logs, but Exim does carry on, to try to do its
6645 job under whatever limits there are.
6647 18. Imported PCRE 5.0.
6649 19. Trivial typo in log message " temporarily refused connection" (the leading
6652 20. If the log selector return_path_on_delivery was set and an address was
6653 redirected to /dev/null, the delivery process crashed because it assumed
6654 that a return path would always be set for a "successful" delivery. In this
6655 case, the whole delivery is bypassed as an optimization, and therefore no
6658 21. Internal re-arrangement: the function for sending a challenge and reading
6659 a response while authentication was assuming a zero-terminated challenge
6660 string. It's now changed to take a pointer and a length, to allow for
6661 binary data in such strings.
6663 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6665 23. Exim was not respecting finduser_retries when seeking the login of the
6666 uid under which it was called; it was always trying 10 times. (The default
6667 setting of finduser_retries is zero.) Also, it was sleeping after the final
6668 failure, which is pointless.
6670 24. Implemented tls_on_connect_ports.
6672 25. Implemented acl_smtp_predata.
6674 26. If the domain in control=submission is set empty, Exim assumes that the
6675 authenticated id is a complete email address when it generates From: or
6676 Sender: header lines.
6678 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6679 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6680 chown and chgrp in /bin and hostname in /usr/bin.
6682 28. Exim was keeping the "process log" file open after each use, just as it
6683 does for the main log. This opens the possibility of it remaining open for
6684 long periods when the USR1 signal hits a daemon. Occasional processlog
6685 errors were reported, that could have been caused by this. Anyway, it seems
6686 much more sensible not to leave this file open at all, so that is what now
6689 29. The long-running daemon process does not normally write to the log once it
6690 has entered its main loop, and it closes the log before doing so. This is
6691 so that log files can straightforwardly be renamed and moved. However,
6692 there are a couple of unusual error situations where the daemon does write
6693 log entries, and I had neglected to close the log afterwards.
6695 30. The text of an SMTP error response that was received during a remote
6696 delivery was being truncated at 512 bytes. This is too short for some of
6697 the long messages that one sometimes sees. I've increased the limit to
6700 31. It is now possible to make retry rules that apply only when a message has a
6701 specific sender, in particular, an empty sender.
6703 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6704 it possible to be selective about when SMTP synchronization is enforced.
6706 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6708 32. Implemented hosts_connection_nolog.
6710 33. Added an ACL for QUIT.
6712 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6715 35. Added mailbox_size and mailbox_filecount to appendfile.
6717 36. Added control = no_multiline_responses to ACLs.
6719 37. There was a bug in the logic of the code that waits for the clock to tick
6720 in the case where the clock went backwards by a substantial amount such
6721 that the microsecond fraction of "now" was more than the microsecond
6722 fraction of "then" (but the whole seconds number was less).
6724 38. Added support for the libradius Radius client library this is found on
6725 FreeBSD (previously only the radiusclient library was supported).
6731 1. When certain lookups returned multiple values in the form name=value, the
6732 quoting of the values was not always being done properly. Specifically:
6733 (a) If the value started with a double quote, but contained no whitespace,
6735 (b) If the value contained whitespace other than a space character (i.e.
6736 tabs or newlines or carriage returns) it was not quoted.
6737 This fix has been applied to the mysql and pgsql lookups by writing a
6738 separate quoting function and calling it from the lookup code. The fix
6739 should probably also be applied to nisplus, ibase and oracle lookups, but
6740 since I cannot test any of those, I have not disturbed their existing code.
6742 2. A hit in the callout cache for a specific address caused a log line with no
6743 reason for rejecting RCPT. Now it says "Previous (cached) callout
6744 verification failure".
6746 3. There was an off-by-one bug in the queryprogram router. An over-long
6747 return line was truncated at 256 instead of 255 characters, thereby
6748 overflowing its buffer with the terminating zero. As well as fixing this, I
6749 have increased the buffer size to 1024 (and made a note to document this).
6751 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6752 when Exim is waiting for an SMTP response from a remote server, Exim
6753 restarts its select() call on the socket, thereby resetting its timeout.
6754 This is not a problem when such interrupts are rare. Somebody set up a cron
6755 job to run exiwhat every 2 minutes, which is less than the normal select()
6756 timeout (5 or 10 minutes). This meant that the select() timeout never
6757 kicked in because it was always reset. I have fixed this by comparing the
6758 time when an interrupt arrives with the time at the start of the first call
6759 to select(). If more time than the timeout has elapsed, the interrupt is
6760 treated as a timeout.
6762 5. Some internal re-factoring in preparation for the addition of Sieve
6763 extensions (by MH). In particular, the "personal" test is moved to a
6764 separate function, and given an option for scanning Cc: and Bcc: (which is
6765 not set for Exim filters).
6767 6. When Exim created an email address using the login of the caller as the
6768 local part (e.g. when creating a From: or Sender: header line), it was not
6769 quoting the local part when it contained special characters such as @.
6771 7. Installed new OpenBSD configuration files.
6773 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6774 try to make them clearer.
6776 9. Callout options, other than the timeout value, were being ignored when
6777 verifying sender addresses in header lines. For example, when using
6779 verify = header_sender/callout=no_cache
6781 the cache was (incorrectly) being used.
6783 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6784 only the Cygwin environment.
6786 11. When return_path_on_delivery was set as a log selector, if different remote
6787 addresses in the same message used different return paths and parallel
6788 remote delivery occurred, the wrong values would sometimes be logged.
6789 (Whenever a remote delivery process finished, the return path value from
6790 the most recently started remote delivery process was logged.)
6792 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6793 lines when AUTH and/or TLS are in use. This is the "received protocol"
6794 field. Exim used to use "asmtp" for authenticated SMTP, without any
6795 indication (in the protocol name) for TLS use. Now it follows the RFC and
6796 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6797 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6798 names appear in log lines as well as in Received: header lines.
6800 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6801 extensions, and comparison tests, and to fix some bugs.
6803 14. Changes to the "personal" filter test:
6805 (1) The test was buggy in that it was just doing the equivalent of
6806 "contains" tests on header lines. For example, if a user's address was
6807 anne@some.where, the "personal" test would incorrectly be true for
6809 To: susanne@some.where
6811 This test is now done by extracting each address from the header in turn,
6812 and checking the entire address. Other tests that are part of "personal"
6813 are now done using regular expressions (for example, to check local parts
6814 of addresses in From: header lines).
6816 (2) The list of non-personal local parts in From: addresses has been
6817 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6818 taken from the Sieve specification recommendations.
6820 (3) If the message contains any header line starting with "List-" it is
6821 treated as non-personal.
6823 (4) The test for "circular" in the Subject: header line has been removed
6824 because it now seems ill-conceived.
6826 15. Minor typos in src/EDITME comments corrected.
6828 16. Installed latest exipick from John Jetmore.
6830 17. If headers_add on a router specified a text string that was too long for
6831 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6832 of string_sprintf() is now avoided.
6834 18. $message_body_size was not set (it was always zero) when running the DATA
6835 ACL and the local_scan() function.
6837 19. For the "mail" command in an Exim filter, no default was being set for
6838 the once_repeat time, causing a random time value to be used if "once" was
6839 specified. (If the value happened to be <= 0, no repeat happened.) The
6840 default is now 0s, meaning "never repeat". The "vacation" command was OK
6841 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6842 (I found it when inspecting the code).
6844 20. There is now an overall timeout for performing a callout verification. It
6845 defaults to 4 times the callout timeout, which applies to individual SMTP
6846 commands during the callout. The overall timeout applies when there is more
6847 than one host that can be tried. The timeout is checked before trying the
6848 next host. This prevents very long delays if there are a large number of
6849 hosts and all are timing out (e.g. when the network connections are timing
6850 out). The value of the overall timeout can be changed by specifying an
6851 additional sub-option for "callout", called "maxwait". For example:
6853 verify = sender/callout=5s,maxwait=20s
6855 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6856 to the end before writing, but this should make it even safer).
6858 22. Exim was forgetting that it had advertised PIPELINING for the second and
6859 subsequent messages on an SMTP connection. It was also not resetting its
6860 memory on STARTTLS and an internal HELO.
6862 23. When Exim logs an SMTP synchronization error within a session, it now
6863 records whether PIPELINING has been advertised or not.
6865 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6866 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6867 rather than long int.
6869 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6871 26. Added the never_mail option to autoreply.
6877 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6878 crash if the getsockname() call failed; this can happen if a connection is
6879 closed very soon after it is established. The problem was simply in the
6880 order in which certain operations were done, causing Exim to try to write
6881 to the SMTP stream before it had set up the file descriptor. The bug has
6882 been fixed by making things happen in the correct order.
6888 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6889 before the connection was closed, thus losing the rejection response.
6891 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6892 some early Solaris releases, but causes trouble in current releases where
6893 socklen_t is defined.
6895 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6898 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6901 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6902 that releasing the top of it at the end releases what was used for sub-
6903 expansions (unless the block got too big). However, discard this block if
6904 the first thing is a variable or header, so that we can use its block when
6905 it is dynamic (useful for very large $message_headers, for example).
6907 6. Lookups now cache *every* query, not just the most recent. A new, separate
6908 store pool is used for this. It can be recovered when all lookup caches are
6909 flushed. Lookups now release memory at the end of their result strings.
6910 This has involved some general refactoring of the lookup sources.
6912 7. Some code has been added to the store_xxx() functions to reduce the amount
6913 of flapping under certain conditions.
6915 8. log_incoming_interface used to affect only the <= reception log lines. Now
6916 it causes the local interface and port to be added to several more SMTP log
6917 lines, for example "SMTP connection from", and rejection lines.
6919 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6921 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6923 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6924 could be overwritten at the end of the current message (or the start of a
6925 new message if it was set in a HELO ACL). The value is now preserved for
6926 the duration of the SMTP connection.
6928 12. If a transport had a headers_rewrite setting, and a matching header line
6929 contained an unqualified address, that address was qualified, even if it
6930 did not match any rewriting rules. The underlying bug was that the values
6931 of the flags that permit the existence of unqualified sender and recipient
6932 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6933 non-local messages, and by -bnq for local messages) were not being
6934 preserved with the message after it was received.
6936 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6937 "next input=" as part of the text comprising the host identity instead of
6938 the correct text. The code was using the same buffer for two different
6939 strings. However, depending on which order the printing function evaluated
6940 its arguments, the bug did not always show up. Under Linux, for example, my
6941 test suite worked just fine.
6943 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6944 doesn't work with some older versions of Perl. It has been changed to "my",
6945 which in any case is probably the better facility to use.
6947 15. A really picky compiler found some instances of statements for creating
6948 error messages that either had too many or two few arguments for the format
6951 16. The size of the buffer for calls to the DNS resolver has been increased
6952 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6953 for addresses that have a lot of PTR records. This alleviates a problem; it
6954 does not fully solve it.
6956 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6957 buffer now truncates the list and logs the incident, which is the same
6958 action as happens when Exim is looking up a host name and its aliases.
6959 Previously in this situation something unpredictable would happen;
6960 sometimes it was "internal error: store_reset failed".
6962 18. If a server dropped the connection unexpectedly when an Exim client was
6963 using GnuTLS and trying to read a response, the client delivery process
6964 crashed while trying to generate an error log message.
6966 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6967 string, for example:
6969 warn message = H1: something\nH2: something
6971 the text was added as a single header line from Exim's point of view
6972 though it ended up OK in the delivered message. However, searching for the
6973 second and subsequent header lines using $h_h2: did not work. This has been
6974 fixed. Similarly, if a system filter added multiple headers in this way,
6975 the routers could not see them.
6977 20. Expanded the error message when iplsearch is called with an invalid key to
6978 suggest using net-iplsearch in a host list.
6980 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6981 ACLs are no longer actually imposed (and a message to that effect is
6984 22. If a "gecos" field in a passwd entry contained escaped characters, in
6985 particular, if it contained a \" sequence, Exim got it wrong when building
6986 a From: or a Sender: header from that name. A second bug also caused
6987 incorrect handling when an unquoted " was present following a character
6988 that needed quoting.
6990 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6991 was not being matched caselessly.
6993 24. Arranged for all hyphens in the exim.8 source to be escaped with
6996 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6997 properly. Recipient callout cache records were still being keyed to include
6998 the sender, even when use_sender was set false. This led to far more
6999 callouts that were necessary. The sender is no longer included in the key
7000 when use_sender is false.
7002 26. Added "control = submission" modifier to ACLs.
7004 27. Added the ${base62d: operator to decode base 62 numbers.
7006 28. dnsdb lookups can now access SRV records.
7008 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7009 the configuration file.
7011 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7012 (-v) mode. This makes the output for a verbose queue run more intelligible.
7014 31. Added a use_postmaster feature to recipient callouts.
7016 32. Added the $body_zerocount variable, containing the number of binary zero
7017 bytes in the message body.
7019 33. The time of last modification of the "new" subdirectory is now used as the
7020 "mailbox time last read" when there is a quota error for a maildir
7023 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7025 35. Added +ignore_unknown as a special item in host lists.
7027 36. Code for decoding IPv6 addresses in host lists is now included, even if
7028 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7029 address was recognized as an IP address, but was then not correctly decoded
7030 into binary, causing unexpected and incorrect effects when compared with
7037 1. Very minor rewording of debugging text in manualroute to say "list of
7038 hosts" instead of "hostlist".
7040 2. If verify=header_syntax was set, and a header line with an unqualified
7041 address (no domain) and a large number of spaces between the end of the
7042 name and the colon was received, the reception process suffered a buffer
7043 overflow, and (when I tested it) crashed. This was caused by some obsolete
7044 code that should have been removed. The fix is to remove it!
7046 3. When running in the test harness, delay a bit after writing a bounce
7047 message to get a bit more predictability in the log output.
7049 4. Added a call to search_tidyup() just before forking a reception process. In
7050 theory, someone could use a lookup in the expansion of smtp_accept_max_
7051 per_host which, without the tidyup, could leave open a database connection.
7053 5. Added the variables $recipient_data and $sender_data which get set from a
7054 lookup success in an ACL "recipients" or "senders" condition, or a router
7055 "senders" option, similar to $domain_data and $local_part_data.
7057 6. Moved the writing of debug_print from before to after the "senders" test
7060 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7061 problems for message scanning, either using a data ACL, or using
7062 local_scan() because the Received: header was not generated till after they
7063 were called (in order to set the time as the time of reception completion).
7064 I have revised the way this works. The header is now generated after the
7065 body is received, but before the ACL or local_scan() are called. After they
7066 are run, the timestamp in the header is updated.
7072 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7073 before starting a queue runner without re-exec. This happened only when
7074 deliver_drop_privilege was set or when the Exim user was set to root. The
7075 effect of the bug was that timeouts during subsequent deliveries caused
7076 crashes instead of being properly handled. The handler is now left at its
7077 default (and expected) setting.
7079 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7080 message, again when deliver_drop_privilege is set or Exim is run as root.
7081 The bug described in (1) was not present in this case, but the tidying up
7082 of the other signals was missing. I have made the two cases consistent.
7084 3. The ignore_target_hosts setting on a manualroute router was being ignored
7085 for hosts that were looked up using the /MX notation.
7087 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7090 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7091 operated on the sender address. After changing the $sender_address to <>
7092 for the sender address verify, Exim was re-instated it as the original
7093 (before rewriting) address, but remembering that it had rewritten it, so it
7094 wasn't rewriting it again. This bug also had the effect of breaking the
7095 sender address verification caching when the sender address was rewritten.
7097 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7098 This has been changed so that if the ip literal address matches
7099 ignore_target_hosts, the router declines.
7101 7. Added expansion conditions match_domain, match_address, and match_local_
7102 part (NOT match_host).
7104 8. The placeholder for the Received: header didn't have a length field set.
7106 9. Added code to Exim itself and to exim_lock to test for a specific race
7107 condition that could lead to file corruption when using MBX delivery. The
7108 issue is with the lockfile that is created in /tmp. If this file is removed
7109 after a process has opened it but before that process has acquired a lock,
7110 there is the potential for a second process to recreate the file and also
7111 acquire a lock. This could lead to two Exim processes writing to the file
7112 at the same time. The added code performs the same test as UW imapd; it
7113 checks after acquiring the lock that its file descriptor still refers to
7114 the same named file.
7116 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7117 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7120 11. Added the smtp_active_hostname option. If used, this will typically be made
7121 to depend on the incoming interface address. Because $interface_address is
7122 not set up until the daemon has forked a reception process, error responses
7123 that can happen earlier (such as "too many connections") no longer contain
7126 12. If an expansion in a condition on a "warn" statement fails because a lookup
7127 defers, the "warn" statement is abandoned, and the next ACL statement is
7128 processed. Previously this caused the whole ACL to be aborted.
7130 13. Added the iplsearch lookup type.
7132 14. Added ident_timeout as a log selector.
7134 15. Added tls_certificate_verified as a log selector.
7136 16. Added a global option tls_require_ciphers (compare the smtp transport
7137 option of the same name). This controls incoming TLS connections.
7139 17. I finally figured out how to make tls_require_ciphers do a similar thing
7140 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7141 before starting the TLS session.
7143 18. Tabs are now shown as \t in -bP output.
7145 19. If the log selector return_path_on_delivery was set, Exim crashed when
7146 bouncing a message because it had too many Received: header lines.
7148 20. If two routers both had headers_remove settings, and the first one included
7149 a superfluous trailing colon, the final name in the first list and the
7150 first name in the second list were incorrectly joined into one item (with a
7151 colon in the middle).
7157 1. Added -C and -D options to the exinext utility, mainly to make it easier
7158 to include in the automated testing, but these could be helpful when
7159 multiple configurations are in use.
7161 2. The exinext utility was not formatting the output nicely when there was
7162 an alternate port involved in the retry record key, nor when there was a
7163 message id as well (for retries that were specific to a specific message
7164 and a specific host). It was also confused by IPv6 addresses, because of
7165 the additional colons they contain. I have fixed the IPv4 problem, and
7166 patched it up to do a reasonable job for IPv6.
7168 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7169 delivery, the log line now contains "pipelined" if PIPELINING was used.
7171 4. An SMTP transport process used to panic and die if the bind() call to set
7172 an explicit outgoing interface failed. This has been changed; it is now
7173 treated in the same way as a connect() failure.
7175 5. A reference to $sender_host_name in the part of a conditional expansion
7176 that was being skipped was still causing a DNS lookup. This no longer
7179 6. The def: expansion condition was not recognizing references to header lines
7180 that used bh_ and bheader_.
7182 7. Added the _cache feature to named lists.
7184 8. The code for checking quota_filecount in the appendfile transport was
7185 allowing one more file than it should have been.
7187 9. For compatibility with Sendmail, the command line option
7195 and sets the incoming protocol and host name (for trusted callers). The
7196 host name and its colon can be omitted when only the protocol is to be set.
7197 Note the Exim already has two private options, -pd and -ps, that refer to
7198 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7199 "s", but I don't think that's a major issue.
7201 10. A number of refactoring changes to the code, none of which should affect
7204 (a) The number of logging options was getting close to filling up the
7205 32-bit word that was used as a bit map. I have split them into two classes:
7206 those that are passed in the argument to log_write(), and those that are
7207 only ever tested independently outside of that function. These are now in
7208 separate 32-bit words, so there is plenty of room for expansion again.
7209 There is no change in the user interface or the logging behaviour.
7211 (b) When building, for example, log lines, the code previously used a
7212 macro that called string_cat() twice, in order to add two strings. This is
7213 not really sufficiently general. Furthermore, there was one instance where
7214 it was actually wrong because one of the argument was used twice, and in
7215 one call a function was used. (As it happened, calling the function twice
7216 did not affect the overall behaviour.) The macro has been replaced by a
7217 function that can join an arbitrary number of extra strings onto a growing
7220 (c) The code for expansion conditions now uses a table and a binary chop
7221 instead of a serial search (which was left over from when there were very
7222 few conditions). Also, it now recognizes conditions like "pam" even when
7223 the relevant support is not compiled in: a suitably worded error message is
7224 given if an attempt is made to use such a condition.
7226 11. Added ${time_interval:xxxxx}.
7228 12. A bug was causing one of the ddress fields not to be passed back correctly
7229 from remote delivery subprocesses. The field in question was not being
7230 subsequently used, so this caused to problems in practice.
7232 13. Added new log selectors queue_time and deliver_time.
7234 14. Might have fixed a bug in maildirsizefile handling that threw up
7235 "unexpected character" debug warnings, and recalculated the data
7236 unnecessarily. In any case, I expanded the warning message to give more
7239 15. Added the message "Restricted characters in address" to the statements in
7240 the default ACL that block characters like @ and % in local parts.
7242 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7243 Three changes have been made:
7245 (a) There was a serious bug; a negative response to MAIL caused the whole
7246 recipient domain to be cached as invalid, thereby blocking all messages
7247 to all local parts at the same domain, from all senders. This bug has
7248 been fixed. The domain is no longer cached after a negative response to
7249 MAIL if the sender used is not empty.
7251 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7254 (c) A new callout option, "use_sender" has been added for people who want
7255 the modified behaviour.
7261 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7264 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7265 indeed breaks things for older releases.
7267 3. Added additional logging to the case where there is a problem reading data
7268 from a filter that is running in a subprocess using a pipe, in order to
7269 try to track down a specific problem.
7271 4. Testing facility fudge: when running in the test harness and attempting
7272 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7273 getting "No route to host". Convert this to a timeout.
7275 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7278 6. Some OS don't have socklen_t but use size_t instead. This affects the
7279 fifth argument of getsockopt() amongst other things. This is now
7280 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7281 can be set for individual OS. I have set it for SunOS5, OSF1, and
7282 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7283 some earlier ones do not.
7285 7. Change 4.30/15 was not doing the test caselessly.
7287 8. The standard form for an IPv6 address literal was being rejected by address
7288 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7289 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7290 this, as well as the form without the "IPv6" on the front (but only when
7291 address literals are enabled, of course).
7293 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7295 10. Exim crashed if a message with an empty sender address specified by -f
7296 encountered a router with an errors_to setting. This could be provoked only
7297 by a command such as
7301 where an empty string was supplied; "<>" did not hit this bug.
7303 11. Installed PCRE release 4.5.
7305 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7306 remained set. It is now erased.
7308 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7309 times from message ids (which are base 36 rather than the normal 62).
7311 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7312 were being counted as actual protocol errors, and logged if the log
7313 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7314 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7315 and DATA following a set of rejected RCPTs do not count as protocol errors.
7316 In other words, Exim assumes they were pipelined, though this may not
7317 actually be the case. Of course, in all cases the client gets an
7318 appropriate error code.
7320 15. If a lookup fails in an ACL condition, a message about the failure may
7321 be available; it is used if testing the ACL cannot continue, because most
7322 such messages specify what the cause of the deferral is. However, some
7323 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7324 that caused an old message to be retained and used if a later statement
7325 caused a defer, replacing the real cause of the deferral.
7327 16. If an IP address had so many PTR records that the DNS lookup buffer
7328 was not large enough to hold them, Exim could crash while trying to process
7329 the truncated data. It now detects and logs this case.
7331 17. Further to 4.21/58, another change has been made: if (and only if) the
7332 first line of a message (the first header line) ends with CRLF, a bare LF
7333 in a subsequent header line has a space inserted after it, so as not to
7334 terminate the header.
7336 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7337 unnecessarily, used atoi() instead of strtol(), and didn't check the
7338 termination when getting file sizes from file names by regex.
7340 19. Completely re-implemented the support for maildirsize files, in the light
7341 of a number of problems with the previous contributed implementation
7342 (4.30/29). In particular:
7344 . If the quota is zero, the maildirsize file is maintained, but no quota is
7347 . If the maildir directory does not exist, it is created before any attempt
7348 to write a maildirsize file.
7350 . The quota value in the file is just a cache; if the quota is changed in
7351 the transport, the new value overrides.
7353 . A regular expression is available for excluding directories from the
7356 20. The autoreply transport checks the characters in options that define the
7357 message's headers; it allows continued headers, but it was checking with
7358 isspace() after an embedded newline instead of explicitly looking for a
7361 21. If all the "regular" hosts to which an address was routed had passed their
7362 expiry times, and had not reached their retry times, the address was
7363 bounced, even if fallback hosts were defined. Now Exim should go on to try
7366 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7367 equivalent code in the SMTP transport. Some hosts send humungous responses
7368 to HELO/EHLO, more than 1024 it seems.
7370 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7371 gives compiler warnings in some environments. I've now done it "properly",
7374 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7375 (because of problems with the built-in one) was declared to return uschar *
7376 instead of char *, causing compiler failure.
7378 25. Fixed a file descriptor leak when processing alias/forward files.
7380 26. Fixed a minor format string issue in dbfn.c.
7382 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7384 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7385 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7388 29. When a sender address is verified, it is cached, to save repeating the test
7389 when there is more than one recipient in a message. However, when the
7390 verification involves a callout, it is possible for different callout
7391 options to be set for different recipients. It is too complicated to keep
7392 track of this in the cache, so now Exim always runs a verification when a
7393 callout is required, relying on the callout cache for the optimization.
7394 The overhead is duplication of the address routing, but this should not be
7397 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7398 to be verified with callout=postmaster, and the main callout worked but the
7399 postmaster check failed, the verification correctly failed. However, if a
7400 subsequent RCPT command asked for sender verification *without* the
7401 postmaster check, incorrect caching caused this verification also to fail,
7404 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7405 it was not caching the DNS options (qualify_single, search_parents) that
7406 were used when the lookup failed. A subsequent lookup with different
7407 options therefore always gave the same answer, though there were cases
7408 where it should not have. (Example: a "domains = !$mx_any" option on a
7409 dnslookup router: the "domains" option is always processed without any
7410 widening, but the router might have qualify_single set.) Now Exim uses the
7411 cached value only when the same options are set.
7413 32. Added John Jetmore's "exipick" utility to the distribution.
7415 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7416 than a timeout (e.g. a certificate is required, and is not provided), an
7417 Exim server now closes the connection immediately. Previously it waited for
7418 the client to close - but if the client is SSL, it seems that they each
7419 wait for each other, leading to a delay before one of them times out.
7421 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7422 maintained 0.8.x compatibility because I don't think many are using it, and
7423 it is clearly obsolete.
7425 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7428 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7429 client certificate was expired. A simple patch fixes this, though I don't
7430 understand the full logic of why the verify callback is called multiple
7433 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7434 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7435 which causes problems with some clients (such as the Certicom SSL Plus
7436 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7437 disables the coutermeasure allowing Eudora to connect."
7439 38. Exim was not checking that a write() to a log file succeeded. This could
7440 lead to Bad Things if a log got too big, in particular if it hit a file
7441 size limit. Exim now panics and dies if it cannot write to a log file, just
7442 as it does if it cannot open a log file.
7444 39. Modified OS/Makefile-Linux so that it now contains
7446 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7448 The two -D definitions ensure that Exim is compiled with large file
7449 support, which makes it possible to handle log files that are bigger than
7452 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7453 instance) a domain was checked against a named list that involved a lookup,
7454 causing $domain_data to be set, then another domain was checked against the
7455 same list, then the first domain was re-checked, the value of $domain_data
7456 after the final check could be wrong. In particular, if the second check
7457 failed, it could be set empty. This bug probably also applied to
7460 41. The strip_trailing_dot option was not being applied to the address given
7461 with the -f command-line option.
7463 42. The code for reading a message's header from the spool was incrementing
7464 $received_count, but never initializing it. This meant that the value was
7465 incorrect (doubled) while delivering a message in the same process in which
7466 it was received. In the most common configuration of Exim, this never
7467 happens - a fresh exec is done - but it can happen when
7468 deliver_drop_privilege is set.
7470 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7471 it now includes up to 150 characters of the unexpected data in the log
7474 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7475 and building data strings. The size of both of these buffers was 10 000
7476 bytes - far larger than anybody would *ever* want, thought I. Needless to
7477 say, somebody hit the limit. I have increased the maximum line length to
7478 20 000 and the maximum data length of concatenated lines to 100 000. I have
7479 also fixed two bugs, because there was no checking on these buffers. Tsk,
7480 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7481 buffer is too small.
7483 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7484 lsearch lookups. Now it does.
7486 46. When parsing a route_list item in a manualroute router, a fixed-length
7487 buffer was used for the list of hosts. I made this 1024 bytes long,
7488 thinking that nobody would ever have a list of hosts that long. Wrong.
7489 Somebody had a whole pile of complicated expansion conditions, and the
7490 string was silently truncated, leading to an expansion error. It turns out
7491 that it is easier to change to an unlimited length (owing to other changes
7492 that have happened since this code was originally written) than to build
7493 structure for giving a limitation error. The length of the item that
7494 expands into the list of hosts is now unlimited.
7496 47. The lsearch lookup could not handle data where the length of text line was
7497 more than 4095 characters. Such lines were truncated, leading to shortened
7498 data being returned. It should now handle lines of any length.
7500 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7501 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7504 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7505 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7506 no longer happens in comment lines. A list of replacements is now placed
7507 at the head of all of the source files, except those whose only change is
7508 to replace PERL_COMMAND in the very first #! line.
7510 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7511 messages on the queue, with a bottom-up merge sort, using code contributed
7512 by Michael Haardt. This should make operations like -bp somewhat faster on
7513 large queues. It won't affect queue runners, except when queue_run_in_order
7516 51. Installed eximstats 1.31 in the distribution.
7518 52. Added support for SRV lookups to the dnslookup router.
7520 53. If an ACL referred to $message_body or $message_body_end, the value was not
7521 reset for any messages that followed in the same SMTP session.
7523 54. The store-handling optimization for building very long strings was not
7524 differentiating between the different store pools. I don't think this
7525 actually made any difference in practice, but I've tidied it.
7527 55. While running the routers to verify a sender address, $sender_address
7528 was still set to the sender address. This is wrong, because when routing to
7529 send a bounce to the sender, it would be empty. Therefore, I have changed
7530 it so that, while verifying a sender address, $sender_address is set to <>.
7531 (There is no change to what happens when verifying a recipient address.)
7533 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7534 target A or AAAA records (if not already returned) without resetting the
7535 qualify_single or search_parents options of the DNS resolver. These are
7536 inappropriate in this case because the targets of MX and SRV records must
7537 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7538 target that, when qualified, matched something in the local domain. These
7539 two options are now turned off when doing these lookups.
7541 57. It seems that at least some releases of Reiserfs (which does not have the
7542 concept of a fixed number of inodes) returns zero and not -1 for the
7543 number of available inodes. This interacted badly with check_spool_inodes,
7544 which assumed that -1 was the "no such thing" setting. What I have done is
7545 to check that the total number of inodes is greater than zero before doing
7546 the test of how many are available.
7548 58. When a "warn" ACL statement has a log_message modifier, the message is
7549 remembered, and not repeated. This is to avoid a lot of repetition when a
7550 message has many recipients that cause the same warning to be written.
7551 However, Exim was preserving the list of already written lines for an
7552 entire SMTP session, which doesn't seem right. The memory is now reset if a
7553 new message is started.
7555 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7556 debugging output unless log_rewrite was also set.
7558 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7559 of (int)(handle) when we know that handle contains (void *)(-1).
7561 61. The Exim daemon panic-logs an error return when it closes the incoming
7562 connection. However "connection reset by peer" seems to be common, and
7563 isn't really an error worthy of noting specially, so that particular error
7566 62. When Exim is trying to find all the local interfaces, it used to panic and
7567 die if the ioctl to get the interface flags failed. However, it seems that
7568 on at least one OS (Solaris 9) it is possible to have an interface that is
7569 included in the list of interfaces, but for which you get a failure error
7570 for this call. This happens when the interface is not "plumbed" into a
7571 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7572 failure of the "get flags" call assumes that the interface is down.
7574 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7575 makes life easier for people who are doing arithmetic on fields extracted
7576 from dates, where you often get leading zeros that should not be
7577 interpreted as octal.
7579 64. Added qualify_domain to the redirect router, to override the global
7582 65. If a pathologically long header line contained very many addresses (the
7583 report of this problem mentioned 10 000) and each of them was rewritten,
7584 Exim could use up a very large amount of memory. (It kept on making new
7585 copies of the header line as it rewrote, and never released the old ones.)
7586 At the expense of a bit more processing, the header rewriting function has
7587 been changed so that it no longer eats memory in this way.
7589 66. The generation of the Received: header has been moved from the time that a
7590 message starts to be received, to the time that it finishes. The timestamp
7591 in the Received: header should now be very close to that of the <= log
7592 line. There are two side-effects of this change:
7594 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7595 logged header lines no longer include the local Received: line, because
7596 it has not yet been created. The same applies to a copy of the message
7597 that is returned to a non-SMTP sender when a message is rejected.
7599 (b) When a filter file is tested using -bf, no additional Received: header
7600 is added to the test message. After some thought, I decided that this
7603 This change does not affect the value of $received_for. It is still set
7604 after address rewriting, but before local_scan() is called.
7606 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7608 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7609 gave an unhelpful panic error message, and a defer error. I have managed to
7610 change this behaviour so that it now rejects any supplied certificate,
7611 which seems right, as the list of acceptable certificates is empty.
7613 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7614 gave an unhelpful defer error. I have not managed to make this reject any
7615 supplied certificates, but the error message it gives is "no certificate
7616 supplied", which is not helpful.
7618 70. exigrep's output now also includes lines that are not associated with any
7619 message, but which match the given pattern. Implemented by a patch from
7620 Martin Sluka, which also tidied up the Perl a bit.
7622 71. Recipient callout verification, like sender verification, was using <> in
7623 the MAIL FROM command. This isn't really the right thing, since the actual
7624 sender may affect whether the remote host accepts the recipient or not. I
7625 have changed it to use the actual sender in the callout; this means that
7626 the cache record is now keyed on a recipient/sender pair, not just the
7627 recipient address. There doesn't seem to be a real danger of callout loops,
7628 since a callout by the remote host to check the sender would use <>.
7629 [SEE ABOVE: changed after hitting problems.]
7631 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7632 temporary errors. However, in the case of such a code being given after
7633 the end of a data transmission (i.e. after ".") Exim was failing to write
7634 a retry record for the message. (Yes, there was some broken host that was
7635 actually sending 8xx at this point.)
7637 73. An unknown lookup type in a host list could cause Exim to panic-die when
7638 the list was checked. (An example that provoked this was putting <; in the
7639 middle of a list instead of at the start.) If this happened during a DATA
7640 ACL check, a -D file could be left lying around. This kind of configuration
7641 error no longer causes Exim to die; instead it causes a defer error. The
7642 incident is still logged to the main and panic logs.
7644 74. Buglet left over from Exim 3 conversion. The message "too many messages
7645 in one connection" was written to the rejectlog but not the mainlog, except
7646 when address rewriting (yes!) was being logged.
7648 75. Added write_rejectlog option.
7650 76. When a system filter was run not as root (that is, when system_filter_user
7651 was set), the values of the $n variables were not being returned to the
7652 main process; thus, they were not subsequently available in the $sn
7655 77. Added +return_path_on_delivery log selector.
7657 78. A connection timeout was being treated differently from recipients deferred
7658 when testing hosts_max_try with a message that was older than the host's
7659 retry timeout. (The host should not be counted, thus allowing all hosts to
7660 be tried at least once before bouncing.) This may have been the cause of an
7661 occasionally reported bug whereby a message would remain on the queue
7662 longer than the retry timeout, but would be bounced if a delivery was
7663 forced. I say "may" because I never totally pinned down the problem;
7664 setting up timeout/retry tests is difficult. See also the next item.
7666 79. The ultimate address timeout was not being applied to errors that involved
7667 a combination of host plus message (for example, a timeout on a MAIL
7668 command). When an address resolved to a number of possible hosts, and they
7669 were not all tried for each delivery (e.g. because of hosts_max_try), a
7670 message could remain on the queue longer than the retry timeout.
7672 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7675 81. Fixed an obscure SMTP outgoing bug which required at least the following
7676 conditions: (a) there was another message waiting for the same server;
7677 (b) the server returned 5xx to all RCPT commands in the first message so
7678 that the message was not completed; (c) the server dropped the connection
7679 or gave a negative response to the RSET that Exim sends to abort the
7680 transaction. The observed case was a dropped connection after DATA that had
7681 been sent in pipelining mode. That is, the server had advertised PIPELINING
7682 but was not implementing it correctly. The effect of the bug was incorrect
7683 behaviour, such as trying another host, and this could lead to a crash.
7689 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7690 and daemon.c were passed as pointers to ints; they should have been
7691 pointers to socklen_t variables (which are typically unsigned ints).
7693 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7696 3. Fixed a really odd bug that affected only the testing scheme; patching a
7697 certain fixed string in the binary changed the value of another string that
7698 happened to be identical to the end of the original first string.
7700 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7701 name", it returns that address as the IP address. On some operating
7702 systems (e.g. Solaris), it also passes back the IP address string as the
7703 "host name". However, on others (e.g. Linux), it passes back an empty
7704 string. Exim wasn't checking for this, and was changing the host name to an
7705 empty string, assuming it had been canonicalized.
7707 5. Although rare, it is permitted to have more than one PTR record for a given
7708 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7709 all the names associated with an address, because they do in Solaris.
7710 However, it seems that they do not in Linux for data that comes from the
7711 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7712 I found this out when I moved to a new Linux workstation and tried to run
7713 the Exim test suite.
7715 To get round this problem I have changed the code so that it now does its
7716 own call to the DNS to look up PTR records when searching for a host name.
7717 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7718 addresses that are only in /etc/hosts are still found.
7720 This behaviour is, however, controlled by an option called host_lookup_
7721 order, which defaults to "bydns:byaddr". If people want to use the other
7722 order, or indeed, just use one or the other means of lookup, they can
7723 specify it in this variable.
7725 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7726 some operating systems, this comes back from gethostbyaddr() as an empty
7727 string, and this is what Exim used to test for. However, it seems that in
7728 other systems, "." is yielded. Exim now tests for this case too.
7730 7. The values of check_spool_space and check_log_space are now held internally
7731 as a number of kilobytes instead of an absolute number of bytes. If a
7732 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7733 kilobyte. This means that much larger values can be stored.
7735 8. Exim monitor: an attempt to get the action menu when not actually pointing
7736 at a message produces an empty menu entitled "No message selected". This
7737 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7738 no entries in it ("Shell widget menu has zero width and/or height"). So I
7739 have added a single, blank menu entry in this case.
7741 9. Added ${quote_local_part.
7743 10. MIME decoding is now applied to the contents of Subject: header lines when
7746 11. Now that a reference to $sender_host_address automatically causes a reverse
7747 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7748 host lookup before query-style lookups in lists that might use this
7749 variable. This has therefore been abolished, and the "net-" prefix is no
7750 longer necessary for query-style lookups.
7752 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7753 have been a typo for LFLAGS, so it has been changed.
7755 13. The install script calls Exim with "-C /dev/null" in order to find the
7756 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7757 to be output. However, since Exim outputs its version number before the
7758 error, it didn't break the script. It just looked ugly. I fixed this by
7759 always allowing "-C /dev/null" if the caller is root.
7761 14. Ignore overlarge ACL variable number when reading spool file - insurance
7762 against a later release with more variables having written the file.
7764 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7765 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7766 this, as well as the form without the "IPv6" on the front.
7768 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7769 OS/Makefile-Darwin file.
7771 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7772 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7774 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7775 variable. However, some OS use ints for inodes. I've added cast to long int
7776 to get rid of the compiler warning.
7778 19. I had forgotten to lock out "/../" in configuration file names when
7779 ALT_CONFIG_PREFIX was set.
7781 20. Routers used for verification do not need to specify transports. However,
7782 if such a router generated a host list, and callout was configured, Exim
7783 crashed, because it could not find a port number from the (non-existent)
7784 transport. It now assumes port 25 in this circumstance.
7786 21. Added the -t option to exigrep.
7788 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7789 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7790 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7791 them is set, LOOKUP_LSEARCH is forced.
7793 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7795 24. Added sender and host information to the "rejected by local_scan()" log
7796 line; previously there was no indication of these.
7798 25. Added .include_if_exists.
7800 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7801 Linux. It turns out that not all file systems support this. Apparently some
7802 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7803 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7804 not supported on the file descriptor, is now ignored when Exim is trying to
7805 sync a directory. This applies only to Linux.
7807 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7809 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7810 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7813 29. Added support for maildirsize files from supplied patch (modified a bit).
7815 30. The use of :fail: followed by an empty string could lead Exim to respond to
7816 sender verification failures with (e.g.):
7818 550 Verification failed for <xxx>
7819 550 Sender verify failed
7821 where the first response line was missing the '-' that indicates it is not
7822 the final line of the response.
7824 31. The loop for finding the name of the user that called Exim had a hardwired
7825 limit of 10; it now uses the value of finduser_retries, which is used for
7826 all other user lookups.
7828 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7831 33. Exim was neglecting to zero errno before one call of strtol() when
7832 expanding a string and expecting an integer value. On some systems this
7833 resulted in spurious "integer overflow" errors. Also, it was casting the
7834 result into an int without checking.
7836 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7837 did not work. The code looks as if it has *never* worked, though it appears
7838 to have been documented since at least release 1.62. I have made it work.
7840 35. The "timeout_DNS" error in retry rules, also documented since at least
7841 1.62, also never worked. As it isn't clear exactly what this means, and
7842 clearly it isn't a major issue, I have abolished the feature by treating it
7843 as "timeout", and writing a warning to the main and panic logs.
7845 36. The display of retry rules for -brt wasn't always showing the error code
7848 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7849 timeout_connect_A, timeout_connect_MX.
7851 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7852 to the empty sender.
7854 39. The daemon was not analysing the content of -oX till after it had closed
7855 stderr and disconnected from the controlling terminal. This meant that any
7856 syntax errors were only noted on the panic log, and the return code from
7857 the command was 0. By re-arranging the code a little, I've made the
7858 decoding happen first, so such errors now appear on stderr, and the return
7859 code is 1. However, the actual setting up of the sockets still happens in
7860 the disconnected process, so errors there are still only recorded on the
7863 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7864 connections (as happens on some IP stacks) was logged at start up time as
7865 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7866 it from "IPv6 and IPv4", which means that two separate sockets are being
7869 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7870 says whether AF_INET or AF_INET6 was passed as an argument.
7872 42. Exiwhat output was messed up when time zones were included in log
7875 43. Exiwhat now gives more information about the daemon's listening ports,
7876 and whether -tls-on-connect was used.
7878 44. The "port" option of the smtp transport is now expanded.
7880 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7881 silently ignored. Now an error message is written to the main and panic
7884 46. There's a new ACL modifier called "logwrite" which writes to a log file
7885 as soon as it is encountered.
7887 47. Added $local_user_uid and $local_user_gid at routing time.
7889 48. Exim crashed when trying to verify a sender address that was being
7892 49. Exim was recognizing only a space character after ".include". It now also
7893 recognizes a tab character.
7895 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7896 extracting the relevant information from the specification. The man page no
7897 longer contains scrambled data for the -d option, and I've added a section
7898 at the front about calling Exim under different names.
7900 51. Added "extra_headers" argument to the "mail" command in filter files.
7902 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7905 53. Installed eximstats 1.29.
7907 54. Added transport_filter_timeout as a generic transport option.
7909 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7910 Cc: header lines. This was required by RFC 822, but it not required by RFC
7913 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7914 incoming messages that did not have them. Now it does so only if the
7915 message originates locally, that is, if there is no associated remote host
7916 address. When Resent- header lines are present, this applies to the Resent-
7917 lines rather than the non-Resent- lines.
7919 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7920 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7922 58. Messages for configuration errors now include the name of the main
7923 configuration file - useful now that there may be more than one file in a
7924 list (.included file names were always shown).
7926 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7927 for those rare installations that do not start the daemon as root or run it
7928 setuid root. I've cut out the call to initgroups() if the daemon is not
7931 60. The Exim user and group can now be bound into the binary as text strings
7932 that are looked up at the start of Exim's processing.
7934 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7936 62. Added $mailstore_basename variable.
7938 63. Installed patch to sieve.c from Michael Haardt.
7940 64. When Exim failed to open the panic log after failing to open the main log,
7941 the original message it was trying to log was written to stderr and debug
7942 output, but if they were not available (the usual case in production), it
7943 was lost. Now it is written to syslog before the two lines that record the
7944 failures to open the logs.
7946 65. Users' Exim filters run in subprocesses under the user's uid. It is
7947 possible for a "deliver" command or an alias in a "personal" command to
7948 provoke an address rewrite. If logging of address rewriting is configured,
7949 this fails because the process is not running as root or exim. There may be
7950 a better way of dealing with this, but for the moment (because 4.30 needs
7951 to be released), I have disabled address rewrite logging when running a
7952 filter in a non-root, non-exim process.
7958 1. The buildconfig auxiliary program wasn't quoting the value set for
7959 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7960 not defined. This bug was masked in 4.22 by the effect that was fixed in
7963 2. Some messages that were rejected after a message id was allocated were
7964 shown as "incomplete" by exigrep. It no longer does this for messages that
7965 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7967 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7968 have allow_domain_literals set, the ID did not get logged in the <= line.
7969 Domain literals are now always recognized in Message-ID: header lines.
7971 4. The first argument for a ${extract expansion item is the key name or field
7972 number. Leading and trailing spaces in this item were not being ignored,
7973 causing some misleading effects.
7975 5. When deliver_drop_privilege was set, single queue runner processes started
7976 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7977 same command in the process it spins off) were not dropping privilege.
7979 6. When the daemon running as "exim" started a queue runner, it always
7980 re-executed Exim in the spun-off process. This is a waste of effort when
7981 deliver_drop_privilege is set. The new process now just calls the
7982 queue-runner function directly.
7988 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7991 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7992 ignored. Though the use of -oP was forcing the writing of a pid file, it
7993 was always written to the default place.
7995 3. If the message "no IP address found for host xxxx" is generated during
7996 incoming verification, it is now followed by identification of the incoming
7997 connection (so you can more easily find what provoked it).
7999 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8001 5. Added some features to "harden" Exim a bit more against certain attacks:
8003 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8004 be put in Local/Makefile. This is like the never_users runtime option,
8005 but it cannot be overridden. The default setting is "root".
8007 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8008 prefix string with which any file named in a -C command line option
8011 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8012 is retained for -C and -D only if the caller of Exim is root. Without
8013 it, the exim user may also use -C and -D and retain privilege.
8015 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8016 command line option is disabled.
8018 6. Macro names set by the -D option must start with an upper case letter, just
8019 like macro names defined in the configuration file.
8021 7. Added "dereference=" facility to LDAP.
8023 8. Two instances of the typo "uknown" in the source files are fixed.
8025 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8026 the Configure-Makefile script screwed up while processing it.
8028 10. Incorporated PCRE 4.4.
8030 11. The SMTP synchronization check was not operating right at the start of an
8031 SMTP session. For example, it could not catch a HELO sent before the client
8032 waited for the greeting. There is now a check for outstanding input at the
8033 point when the greeting is written. Because of the duplex, asynchronous
8034 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8035 way, but not yet received, when the check is performed.
8037 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8038 on TCP/IP sockets, because this apparently causes some broken clients to
8041 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8042 unchanged) from the Cygwin maintainer.
8044 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8045 was supported instead of testing for mbx. Effectively a typo.
8047 15. The spa authenticator server code was not checking that the input it
8048 received was valid base64.
8050 16. The debug output line for the "set" modifier in ACLs was not showing the
8051 name of the variable that was being set.
8053 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8055 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8056 lookup on its own. Something else was needed to trigger the lookup. For
8057 example, a match in host_lookup or the need for a host name in a host list.
8058 Now, if $sender_host_name is referenced and the host name has not yet been
8059 looked up, a lookup is performed. If the lookup fails, the variable remains
8060 empty, and $host_lookup_failed is set to "1".
8062 19. Added "eqi" as a case-independent comparison operator.
8064 20. The saslauthd authentication condition could segfault if neither service
8065 nor realm was specified.
8067 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8068 error message that was logged was misleading, and incoming SMTP
8069 connections were dropped. The message is now more accurate, and temporary
8070 errors are given to SMTP connections.
8072 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8073 (except RSET) with a 421 error, until QUIT is received. However, it was
8074 failing to send a response to QUIT.
8076 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8077 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8078 ACL is now run after the helo_try_verify_hosts code.
8080 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8081 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8082 case-independent, so other case variants are also recognized). Apparently
8083 some systems use these upper case variants.
8085 25. If more than two messages were waiting for the same host, and a transport
8086 filter was specified for the transport, Exim sent two messages over the
8087 same TCP/IP connection, and then failed with "socket operation on non-
8088 socket" when it tried to send the third.
8090 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8092 27. The extern definition of crypt16() in expand.c was not being excluded when
8093 the OS had its own crypt16() function.
8095 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8096 as a preferred synonym for return_size_limit, both as an option and as an
8099 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8101 30. Changed the default configuration ACL to relax the local part checking rule
8102 for addresses that are not in any local domains. For these addresses,
8103 slashes and pipe symbols are allowed within local parts, but the sequence
8104 /../ is explicitly forbidden.
8106 31. SPA server authentication was not clearing the challenge buffer before
8109 32. log_message in a "warn" ACL statement was writing to the reject log as
8110 well as to the main log, which contradicts the documentation and doesn't
8111 seem right (because no rejection is happening). So I have stopped it.
8113 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8114 However, I am unable to do any testing of this.
8116 34. Fixed an infelicity in the appendfile transport. When checking directories
8117 for a mailbox, to see if any needed to be created, it was accidentally
8118 using path names with one or more superfluous leading slashes; tracing
8119 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8121 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8122 contents are added to the log line that is written for every discarded
8123 recipient. (Previously a log_message setting was ignored.)
8125 36. The ${quote: operator now quotes the string if it is empty.
8127 37. The install script runs exim in order to find its version number. If for
8128 some reason other than non-existence or emptiness, which it checks, it
8129 could not run './exim', it was installing it with an empty version number,
8130 i.e. as "exim-". This error state is now caught, and the installation is
8133 38. An argument was missing from the function that creates an error message
8134 when Exim fails to connect to the socket for saslauthd authentication.
8135 This could cause Exim to crash, or give a corrupted message.
8137 39. Added isip, isip4, and isip6 to ${if conditions.
8139 40. The ACL variables $acl_xx are now saved with the message, and can be
8140 accessed later in routers, transports, and filters.
8142 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8143 strings in the file are not string-expanded.
8145 42. If a MAIL command specified a SIZE value that was too large to fit into an
8146 int variable, the check against message_size_limit failed. Such values are
8147 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8148 day this will have to be increased, but I don't think I want to be around
8149 when emails are that large.
8156 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8157 iconv() is not standard in FreeBSD.
8159 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8160 IPv6 enabled. The observed symptom was a segmentation fault on return from
8161 the function os_common_find_running_interfaces() in src/os.c.
8163 3. In the check_special_case() function in daemon.c I had used "errno" as an
8164 argument name, which causes warnings on some systems. This was basically a
8165 typo, since it was named "eno" in the comments!
8167 4. The code that waits for the clock to tick (at a resolution of some fraction
8168 of a second) so as to ensure message-id uniqueness was always waiting for
8169 at least one whole tick, when it could have waited for less. [This is
8170 almost certainly not relevant at current processor speeds, where it is
8171 unlikely to ever wait at all. But we try to future-proof.]
8173 5. The function that sleeps for a time interval that includes fractions of a
8174 second contained a race. It did not block SIGALRM between setting the
8175 timer, and suspending (a couple of lines later). If the interval was short
8176 and the sigsuspend() was delayed until after it had expired, the suspension
8177 never ended. On busy systems this could lead to processes getting stuck for
8180 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8181 process, before it forks any delivery processes. The open lookup caching
8182 mechanism meant that the open file or database connection was passed into
8183 the delivery process. The problem was that delivery processes always tidy
8184 up cached lookup data. This could cause a problem for the next delivery
8185 process started by the queue runner, because the external queue runner
8186 process does not know about the closure. So the next delivery process
8187 still has data in the lookup cache. In the case of a file lookup, there was
8188 no problem because closing a file descriptor in a subprocess doesn't affect
8189 the parent. However, if the lookup was caching a connection to a database,
8190 the connection was closed, and the second delivery process was likely to
8191 see errors such as "PGSQL: query failed: server closed the connection
8192 unexpectedly". The problem has been fixed by closing all cached lookups
8193 in a queue runner before running a delivery process.
8195 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8196 seem to have the "const" qualifier which it has on other OS. I've
8199 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8200 *of the same type* (client or server) with the same public name that an
8201 error should be diagnosed.
8203 9. When Exim looked up a host name for an IP address, but failed to find the
8204 original IP address when looking up the host name (a safety check), it
8205 output the message "<ip address> does not match any IP for NULL", which was
8206 confusing, to say the least. The bug was that the host name should have
8207 appeared instead of "NULL".
8209 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8210 user that is built into the binary, and the -C or -D options is used, root
8211 privilege is dropped before the configuration file is read. In addition,
8212 logging is switched to stderr instead of the normal log files. If the
8213 configuration then re-defines the Exim user, the unprivileged environment
8214 is probably not what is expected, so Exim logs a panic warning message (but
8217 However, if deliver_drop_privilege is set, the unprivileged state may well
8218 be exactly what is intended, so the warning has been cut out in that case,
8219 and Exim is allowed to try to write to its normal log files.
8225 1. smtp_return_error_details was not giving details for temporary sender
8226 or receiver verification errors.
8228 2. Diagnose a configuration error if two authenticators have the same public
8231 3. Exim used not to create the message log file for a message until the first
8232 delivery attempt. This could be confusing when incoming messages were held
8233 for policy or load reasons. The message log file is now created at the time
8234 the message is received, and an initial "Received" line is written to it.
8236 4. The automatically generated man page for command line options had a minor
8237 bug that caused no ill effects; however, a more serious problem was that
8238 the procedure for building the man page automatically didn't always
8239 operate. Consequently, release 4.20 contains an out-of-date version. This
8240 shouldn't happen again.
8242 5. When building Exim with embedded Perl support, the script that builds the
8243 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8244 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8246 6. The freeze_tell option was not being used for messages that were frozen on
8247 arrival, either by an ACL or by local_scan().
8249 7. Added the smtp_incomplete_transaction log selector.
8251 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8252 was accepting AUTH without a new EHLO.
8254 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8255 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8258 10. Logging of TCP/IP connections (when configured) now happens in the main
8259 daemon process instead of the child process, so that the TCP/IP connection
8260 count is more accurate (but it can never be perfect).
8262 11. The use of "drop" in a nested ACL was not being handled correctly in the
8263 outer ACL. Now, if condition failure induced by the nested "drop" causes
8264 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8265 or "require"), the connection is dropped.
8267 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8268 that yield "discard" can now be used with an "accept" or a "discard" verb,
8269 but an error is generated for any others (because I can't see a useful way
8270 to define what should happen).
8272 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8273 are now processed in the same way as lines in the Exim configuration file.
8274 In particular, continuation lines are supported.
8276 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8278 15. Added -ti meaning -t -i.
8280 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8281 domains, and warn by logging if others are found.
8283 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8284 returned from GIFCONF when Exim is trying to find the list of interfaces on
8285 a host. The code in os.c has been modified to copy each ifreq to an aligned
8286 structure in all cases.
8288 Also, in some cases, the returned ifreq's were being copied to a 'struct
8289 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8290 means the last couple of bytes of an IPv6 address could be chopped if the
8291 ifreq contained only a normal sockaddr (14 bytes storage).
8293 18. Named domain lists were not supported in the hosts_treat_as_local option.
8294 An entry such as +xxxx was not recognized, and was treated as a literal
8297 19. Ensure that header lines added by a DATA ACL are included in the reject log
8298 if the ACL subsequently rejects the message.
8300 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8301 MD5 (which is deprecated).
8303 21. When testing a filter file using -bf, Exim was writing a message when it
8304 took the sender from a "From " line in the message, but it was not doing so
8305 when it took $return_path from a Return-Path: header line. It now does.
8307 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8308 with a valid header line field name (a series of printing characters
8309 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8311 23. Changed "disc" in the source to "disk" to conform to the documentation and
8312 the book and for uniformity.
8314 24. Ignore Sendmail's -Ooption=value command line item.
8316 25. When execve() failed while trying to run a command in a pipe transport,
8317 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8318 could be confused with a return value of 69 from the command itself. This
8319 has been changed to 127, the value the shell returns if it is asked to run
8320 a non-existent command. The wording for the related log line suggests a
8321 non-existent command as the problem.
8323 26. If received_header_text expands to an empty string, do not add a Received:
8324 header line to the message. (Well, it adds a token one on the spool, but
8325 marks it "old" so that it doesn't get used or transmitted.)
8327 27. Installed eximstats 1.28 (addition of -nt option).
8329 28. There was no check for failure on the call to getsockname() in the daemon
8330 code. This can fail if there is a shortage of resources on the system, with
8331 ENOMEM, for example. A temporary error is now given on failure.
8333 29. Contrary to the C standard, it seems that in some environments, the
8334 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8335 program. Exim now does this explicitly; it affects the formatting of
8336 timestamps using strftime().
8338 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8339 complaints. I've now initialized all the variables, to avoid this.
8341 32. Header lines added by a system filter were not being "seen" during
8342 transport-time rewrites.
8344 33. The info_callback() function passed to OpenSSL is set up with type void
8345 (*)(SSL *, int, int), as described somewhere. However, when calling the
8346 function (actually a macro) that sets it up, the type void(*)() is
8347 expected. I've put in a cast to prevent warnings from picky compilers.
8349 34. If a DNS black list lookup found a CNAME record, but there were no A
8350 records associated with the domain it pointed at, Exim crashed.
8352 35. If a DNS black list lookup returned more than one A record, Exim ignored
8353 all but the first. It now scans all returned addresses if a particular IP
8354 value is being sought. In this situation, the contents of the
8355 $dnslist_value variable are a list of all the addresses, separated by a
8358 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8359 to accept a host name and all its aliases if the forward lookup for any of
8360 them yielded the IP address of the incoming connection. Now it accepts only
8361 those names whose forward lookup yields the correct IP address. Any other
8362 names are discarded. This closes a loophole whereby a rogue DNS
8363 administrator could create reverse DNS records to break through a
8364 wildcarded host restriction in an ACL.
8366 37. If a user filter or a system filter that ran in a subprocess used any of
8367 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8368 the wrong values were passed to the pipe command ($thisaddress had the
8369 value of $0, $0 had the value of $1, etc). This bug was introduced by
8370 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8372 38. Improved the line breaking for long SMTP error messages from ACLs.
8373 Previously, if there was no break point between 40 and 75 characters, Exim
8374 left the rest of the message alone. Two changes have been made: (a) I've
8375 reduced the minimum length to 35 characters; (b) if it can't find a break
8376 point between 35 and 75 characters, it looks ahead and uses the first one
8377 that it finds. This may give the occasional overlong line, but at least the
8378 remaining text gets split now.
8380 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8381 file descriptors might be low, and that setting 1000 would always raise it.
8382 It turns out that in some environments, the limit is already over 1000 and
8383 that lowering it causes trouble. So now Exim takes care not to decrease it.
8385 40. When delivering a message, the value of $return_path is set to $sender_
8386 address at the start of routing (routers may change the value). By an
8387 oversight, this default was not being set up when an address was tested by
8388 -bt or -bv, which affected the outcome if any router or filter referred to
8391 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8392 log only when "message" or "log_message" are set. However, if one of the
8393 conditions was an address verification, or a call to a nested ACL, the
8394 messages generated by the underlying test were being passed through. This
8395 no longer happens. The underlying message is available in $acl_verify_
8396 message for both "message" and "log_message" expansions, so it can be
8397 passed through if needed.
8399 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8400 new expansion $bh_ to give the encoded byte string without charset
8401 translation. Translation happens only if iconv() is available; HAVE_ICONV
8402 indicates this at build time. HEADERS_CHARSET gives the charset to
8403 translate to; headers_charset can change it in the configuration, and
8404 "headers charset" can change it in an individual filter file.
8406 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8407 that creates RFC 2047 encoded "words" labels them as that charset instead
8408 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8409 expansion operator; (ii) when Exim creates a From: line for a local
8410 message; (iii) when a header line is rewritten to include a "phrase" part.
8412 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8413 buggy, causing it to skip the first lines of messages whose message ID
8414 ended in 'D'. This would not have bitten before Exim release 4.14, because
8415 message IDs were unlikely to end in 'D' before then. The effect was to have
8416 incorrect size information for certain domains.
8418 45. #include "config.h" was missing at the start of the crypt16.c module. This
8419 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8422 46. If there was a timeout during a "random" callout check, Exim treated it as
8423 a failure of the random address, and carried on sending RSET and the real
8424 address. If the delay was just some slowness somewhere, the response to the
8425 original RCPT would be taken as a response to RSET and so on, causing
8426 mayhem of various kinds.
8428 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8429 when I implemented it. It didn't allow for the fact that some option values
8430 may legitimately be negative (e.g. size_addition), and it didn't even do
8431 the right test for positive values.
8433 48. Domain names in DNS records are case-independent. Exim always looks them up
8434 in lower case. Some resolvers return domain names in exactly the case they
8435 appear in the zone file, that is, they may contain uppercase letters. Not
8436 all resolvers do this - some return always lower case. Exim was treating a
8437 change of case by a resolver as a change of domain, similar to a widening
8438 of a domain abbreviation. This triggered its re-routing code and so it was
8439 trying to route what was effectively the same domain again. This normally
8440 caused routing to fail (because the router wouldn't handle the domain
8441 twice). Now Exim checks for this case specially, and just changes the
8442 casing of the domain that it ultimately uses when it transmits the message
8445 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8448 50. If a filter generated a file delivery with a non-absolute name (possible if
8449 no home directory exists for the router), the forbid_file option was not
8452 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8453 the existing equality matching.
8455 52. Exim was using ints instead of ino_t variables in some places where it was
8456 dealing with inode numbers.
8458 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8459 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8460 called TMPDIR, and if it finds it is different, it changes its value.
8462 54. The smtp_printf() function is now made available to local_scan() so
8463 additional output lines can be written before returning. There is also an
8464 smtp_fflush() function to enable the detection of a dropped connection.
8465 The variables smtp_input and smtp_batched_input are exported to
8468 55. Changed the default runtime configuration: the message "Unknown user"
8469 has been removed from the ACL, and instead placed on the localuser router,
8470 using the cannot_route_message feature. This means that any verification
8471 failures that generate their own messages won't get overridden. Similarly,
8472 the "Unrouteable address" message that was in the ACL for unverifiable
8473 relay addresses has also been removed.
8475 56. Added hosts_avoid_esmtp to the smtp transport.
8477 57. The exicyclog script was not checking for the esoteric option
8478 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8479 will work only if exicyclog is run under the appropriate euid.
8481 58. Following a discussion on the list, the rules by which Exim recognises line
8482 endings on incoming messages have been changed. The -dropcr and drop_cr
8483 options are now no-ops, retained only for backwards compatibility. The
8484 following line terminators are recognized: LF CRLF CR. However, special
8485 processing applies to CR:
8487 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8488 nor a local message in the state where . is a terminator.
8490 (ii) If a bare CR is encountered in a header line, an extra space is added
8491 after the line terminator so as not to end the header. The reasoning
8492 behind this is that bare CRs in header lines are most likely either
8493 to be mistakes, or people trying to play silly games.
8495 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8496 was being incorrectly given as 18 bytes larger than it should have been.
8497 This is a VOB (very old bug).
8499 60. This may never have affected anything current, but just in case it has:
8500 When the local host is found other than at the start of a list of hosts,
8501 the local host, those with the same MX, and any that follow, are discarded.
8502 When the list in question was part of a longer list of hosts, the following
8503 hosts (not currently being processed) were also being discarded. This no
8504 longer happens. I'm not sure if this situation could ever has previously
8507 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8510 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8511 header. This is something that is recommended in a new Internet Draft, and
8512 is something that is documented as being done by Sendmail. There are two
8513 possible values. For messages generated by the autoreply transport, Exim
8516 Auto-Submitted: auto-replied
8518 whereas for all other generated messages (e.g. bounces) it adds
8520 Auto-Submitted: auto-generated
8522 63. The "personal" condition in filters now includes a test for the
8523 Auto-Submitted: header. If it contains the string "auto-" the message it
8524 not considered personal.
8526 64. Added rcpt_include_affixes as a generic transport option.
8528 65. Added queue_only_override (default true).
8530 66. Added the syslog_duplication option.
8532 67. If what should have been the first header line of a message consisted of
8533 a space followed by a colon, Exim was mis-interpreting it as a header line.
8534 It isn't of course - it is syntactically invalid and should therefore be
8535 treated as the start of the message body. The misbehaviour could have
8536 caused a number of strange effects, including loss of data in subsequent
8537 header lines, and spool format errors.
8539 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8540 client host had authenticated. This control can now be exercised by an ACL
8541 for more flexibility.
8543 69. By default, callouts do not happen when testing with -bh. There is now a
8544 variant, -bhc, which does actually run the callout code, including
8545 consulting and updating the callout cache.
8547 70. Added support for saslauthd authentication, courtesy of Alexander
8550 71. If statvfs() failed on the spool or log directories while checking their
8551 size for availability, Exim confusingly gave the error "space shortage".
8552 Furthermore, in debugging mode it crashed with a floating point exception.
8553 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8554 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8555 serious problem, Exim now writes to the main and panic logs when this
8556 happens, with details of the failure. It then refuses to accept the
8557 incoming message, giving the message "spool directory problem" or "log
8558 directory problem" with a 421 code for SMTP messages.
8560 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8561 0, 1, and 2 exist, because some OS complain for execs without them (see
8562 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8563 descriptors. However, the code omitted to check that the open succeeded,
8564 causing mysterious errors if for some reason the permissions on /dev/null
8565 got screwed. Now Exim writes a message to the main and panic logs, and
8566 bombs out if it can't open /dev/null.
8568 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8569 interact so that it is all more flexible. It is supposed to remain
8570 backwards compatible. Also added extra_local_interfaces.
8572 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8573 to bomb out with an assertion failure - to the client this appears as a
8574 connection drop. This problem occurs in the part of the code that was taken
8575 from the Samba project. Fortunately, the assertion is in a very simple
8576 function, so I have fixed this by reproducing the function inline in the
8577 one place where it is called, and arranging for authentication to fail
8578 instead of killing the process with assert().
8580 75. The SPA client code was not working when the server requested OEM rather
8581 than Unicode encoding.
8583 76. Added code to make require_files with a specific uid setting more usable in
8584 the case where statting the file as root fails - usually a non-root-mounted
8585 NFS file system. When this happens and the failure is EACCES, Exim now
8586 forks a subprocess and does the per-uid checking as the relevant uid.
8588 77. Added process_log_path.
8590 78. If log_file_path was not explicitly set, a setting of check_log_space or
8591 check_log_inodes was ignored.
8593 79. If a space check for the spool or log partitions fails, the incident is now
8594 logged. Of course, in the latter case the data may get lost...
8596 80. Added the %p formatting code to string_format() so that it can be used to
8597 print addresses in debug_print(). Adjusted all the address printing in the
8598 debugging in store.c to use %p rather than %d.
8600 81. There was a concern that a line of code in smtp_in.c could overflow a
8601 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8602 initially expressed, the concern was not well-founded, because trailing
8603 spaces are removed early. However, if the trailing spaces were followed by
8604 a NULL, they did not get removed, so the overflow was possible. Two fixes
8607 (a) I re-wrote the offending code in a cleaner fashion.
8608 (b) If an incoming SMTP command contains a NULL character, it is rejected
8611 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8612 runs initgroups(), so that if the Exim user is in any additional groups,
8613 they will be used during message reception.
8619 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob