1 This document describes *changes* to previous versions, that might
2 affect Exim's operation, with an unchanged configuration file. For new
3 options, and new features, see the NewStuff file next to this ChangeLog.
8 JH/01 Support list of dkim results in the dkim_status ACL condition, making
9 it more usable in the data ACL.
11 JH/02 Bug 3040: Handle error on close of the spool data file during reception.
12 Previously This was only logged, on the assumption that errors would be
13 seen for a previous fflush(). However, a fuse filesystem has been
14 reported as showing this an error for the fclose(). The spool is now in
15 an uncertain state, and we have logged and responded acceptance. Change
16 this to respond with a temp-reject, wipe spoolfiles, and log the error
19 JH/03 Bug 3030: Fix handling of DNS servfail respons for DANE TLSA. When hit
20 during a recipient verify callout, a QUIT command was attempted on the
21 now-closed callout channel, causing a paniclog entry.
23 JH/04 Bug 3039: Fix handling of of an empty log_reject_target, with
24 a connection_reject log_selector, under tls_on_connect. Previously
25 with this combination, when the connect ACL rejected, a spurious
26 paniclog entry was made.
28 JH/05 Fix TLS resumption for TLS-on-connect. This was broken by the advent
29 of loadbalancer-detection for resumption, in 4.96 - which tries to
30 use the EHLO response. SMTPS does not have one at the time it is starting
31 TLS. Change the default for the smtp transport host_name_extract option
32 to be a static string, for TLS-on-connect cases; meaning that resumption
33 will always be attempted (unless deliberately overriden).
35 JH/06 Bug 3054: Fix dnsdb lookup for a TXT record with multiple chunks, with a
36 chunk-separator specification. This was broken by hardening introduced
39 JH/07 Bug 3050: Fix -bp for old message_id format spoolfiles. Previously it
40 included the -H with the id; this also messed up exiqgrep.
42 JH/08 Bug 3056: Tighten up parsing of DKIM DNS records. Previously, whitespace
43 was not properly skipped and empty elements would cause mis-parsing.
44 Tighten parsing of DKIM header records. Previously, all but lowercase
45 alpha chars would be ignored in potential tag names.
47 JH/09 Bug 3057: Add heuristic for spotting mistyped IPv6 addresses in lists
48 being searched. Previously we only had one for IPv4 addresses. Per the
49 documentation, the error results by default in a no-match result for the
50 list. It is logged if the unknown_in_list log_selector is used.
52 JH/10 Bug 3058: Ensure that a failing expansion in a router "set" option defers
53 the routing operation. Previously it would silently stop routing the
56 JH/11 Bug 3046: Fix queue-runs. Previously, the arrivel of a notification or
57 info-request event close in time to a scheduled run timer could result in
58 the latter being missed, and no further queue scheduled runs being
59 initiated. This ouwld be more likely on high-load systems.
61 JH/12 Refuse to accept a line "dot, LF" as end-of-DATA unless operating in
62 LF-only mode (as detected from the first header line). Previously we did
63 accept that in (normal) CRLF mode; this has been raised as a possible
64 attack scenario (under the name "smtp smuggling").
66 JH/13 Add an fdatasync call for the received message data file in spool, before
67 loggging reception and sending the SMTP ack. Previously we only flushed
68 the stdio buffer so there was still the possibility of a disk error.
70 JH/14 Bug 3061: Avoid a split log line when trying to rewrite a malformed
71 address. Previously, for the last address in a header line (commonly
72 there is only one) the terminating newline was part of the logged
75 JH/15 Bug 3061: Ensure a log line is written for a malformed address in a
76 header, when parsing for address-qualification. Previously one was only
77 written if there were rewrite rules.
79 JH/16 Two-phase queue runs are now reported in the daemon startup log line and
82 JH/17 Bug 3064: Fix combination of "-q<period> -R <recipients>". Introduction of
83 the multiple-queue-runners facility for 4.97 broke this, giving only a
84 one-time run of the queue.
86 JH/18 Bug 3068: Log a warning for use of deprecated syntax in query-style
89 JH/19 Fix TLS startup. When the last expansion done before the initiation of a
90 TLS session resulted in a forced-fail, a misleading error was logged for
91 the expansino of tls_certificates. This would affect the common case of
92 that option being set (main-section options) but not having any variable
93 parts. It could also potentially affect tls_privatekeys. The underlyding
94 coding errors go back to 4.90 but were only exposed in 4.97.
96 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
97 pcre2 library starting allocating 20kB rather than 112 bytes per match
98 call, which broke the 2GB total limitation on Exim's memory management
99 when a user had over 104207 messages stored and the appendfile
100 maildir_quota_directory_regex option is in use. Release the allocated
101 memory every thosand files to avoid this.
102 The same issue arises with the ACL regex condition, which is applied
103 to every line of a received message.
105 JH/21 Bug 3059: Fix crash in smtp transport. When running for a message for
106 which all recipients had been handled (itself an issue) a null-pointer
107 deref was done on trying to write a retry record. Fix that by counting
108 the outstanding recipients before trying to transmit the message.
109 The situation arose for a second MX try within a transport run, when the
110 first had perm-rejected a recipient (the only one for the connection, in
111 the case seen) during pipelining, and then closed the TCP connection.
112 The transport classified that as an I/O error, leaving the message
113 outstanding but having marked up the recipient as dealt-with. It then
114 tried another MX because of the I/O error. Fix this by converting the
115 message-level status to ok if there was a close but all recipients were
116 dealt with. Thanks to Wolfgand Breyha for debug runs.
118 JH/22 The ESMTP_LIMITS facility (RFC 9422) is promoted from experimental status
119 and is now controlled by the build-time option DISABLE_ESMTP_LIMITS.
121 JH/23 Bug 3066: Avoid leaking lookup database credentials to log.
123 JH/24 Bug 3081: Fix a delivery process crash. When the router "errors_to"
124 option specified a fixed address, later rewriting on that address would
125 trip on the configuration data being readonly. Instead of modifying
126 in-place, copy data. Found and fixed by Peter Benie.
128 JH/25 Bug 3079: Fix crash in dbmnz. When a key was present for zero-length
129 data a null pointer was followed. Find and testcase by Sebastian Bugge.
131 JH/26 Fix encoding for an AUTH parameter on a MAIL FROM command. Previously
132 decimal 127 chars were not encoded, and lowercase hex was used for
133 encoded values. Outstanding since at least 1999.
135 JH/27 Fix crash in logging. When a message with a large number of recipients
136 had been received, and logging of recipients is enabled, the buffer used
137 for logging could reach limit. A read using a null pointer would then
138 be done, resulting in a crash of the receiving process before an SMTP
139 ACK for the message was returned to the sending system. Duplicate
140 messages were created as a result.
141 Find and debug help by Mateusz Krawczyk
143 JH/28 Bug 3086: Fix exinext for ipv6. Change the format of keys in the retry
144 DB, wrapping transport record bare-ip "host names" and ipv6
145 "host addresses" in square-brackets. This makes the parsing that
146 exinext does more reliable.
148 JH/29 Bug 3087: Fix SRS encode. A zero-length quoted element in the local-part
155 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
156 SMTP connection" log lines.
158 JH/02 Option default value updates:
159 - queue_fast_ramp (main) true (was false)
160 - remote_max_parallel (main) 4 (was 2)
162 JH/03 Cache static regex pattern compilations, for use by ACLs.
164 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
165 Make the rewrite never match and keep the logging. Trust the
166 admin to be using verify=header-syntax (to actually reject the message).
168 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
169 (under Linux) we watch the dir containing the final file; previously
170 it would be the dir with the first symlink. We still do not monitor
173 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
174 dn_expand() is happy to pass them through.
176 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
177 the file with the proof had an unchanged name, the new proof(s) were
178 loaded on top of the old ones (and nover used; the old ones were stapled).
180 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
181 more than one message arrived in a single connection a reference from
182 the earlier message could be re-used. Often a sigsegv resulted.
183 These variables were introduced in Exim 4.87.
184 Debug help from Graeme Fowler.
186 JH/09 Fix ${filter } for conditions that modify $value. Previously the
187 modified version would be used in construction the result, and a memory
190 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
191 Find and fix by Jasen Betts.
193 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
194 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
195 the systemwide configuration to override the Exim config.
197 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
198 API changes in libopendmarc.
200 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
201 pid 1, in the normal "background daemon" mode, having to drop process-
202 group leadership also lost track of needing to create listener sockets.
204 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
205 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
207 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
208 a capture group which obtained no text (eg. "(abc)*" matching zero
209 occurrences) could cause a segfault if the corresponding $<n> was
212 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
213 included a close-brace character (eg. it itself used an expansion) an
216 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
217 starting TLS. Previously it was after, meaning that attackers on such
218 ports had to be screened using the host_reject_connection main config
219 option. The new sequence aligns better with the STARTTLS behaviour, and
220 permits defences against crypto-processing load attacks, even though it
221 is strictly an incompatible change.
222 Also, avoid sending any SMTP fail response for either the connect ACL
223 or host_reject_connection, for TLS-on-connect ports.
225 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
226 Previously this was not permitted, but it makes reasonable sense.
227 While there, restore a restriction on using it from a connect ACL; given
228 the change JH/16 it could only return false (and before 4.91 was not
231 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
232 was exactly sized compared to the log buffer, a crash occurred with the
233 misleading message "bad memory reference; pool not found".
234 Found and traced by Jasen Betts.
236 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
237 dns_again_means_nonexist included an element causing a DNS lookup which
238 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
239 included (though probably not limited to) a process crash from stack
240 memory limit, or from excessive open files. Replace this with a paniclog
241 whine (as this is likely a configuration error), and returning
244 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
245 this always failed, probably leading to the usual downgrade to in-clear
248 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
249 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
250 to not checking that list for these lookups.
252 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
255 JH/23 Fix crash in string expansions. Previously, if an empty variable was
256 immediately followed by an expansion operator, a null-indirection read
257 was done, killing the process.
259 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
260 include an SMTP response string which is longer than that supported
261 by the delivering transport. Alleviate by wrapping such lines before
264 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
265 chars (RFC limit). Previously a limit of 12 items was made, which with
266 a not-impossible References: in the message being bounced could still
267 be over-large and get stopped in the transport.
269 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
270 close. Previously a bare socket close was done.
272 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
275 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
276 message being considered for delivery was already being handled by
277 another process, and having an SMTP connection already open, the function
278 to close it tried to use an uninitialized variable. This would afftect
279 high-volume sites more, especially when running mailing-list-style loads.
280 Pollution of logs was the major effect, as the other process delivered
281 the message. Found and partly investigated by Graeme Fowler.
283 JH/29 Change format of the internal ID used for message identification. The old
284 version only supported 31 bits for a PID element; the new 64 (on systems
285 which can use Base-62 encoding, which is all currently supported ones
286 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
287 and must use Base-36). The new ID is 23 characters rather than 16, and is
288 visible in various places - notably logs, message headers, and spool file
289 names. Various of the ancillary utilities also have to know the format.
290 As well as the expanded PID portion, the sub-second part of the time
291 recorded in the ID is expanded to support finer precision. Theoretically
292 this permits a receive rate from a single comms channel of better than the
294 The major timestamp part of the ID is not changed; at 6 characters it is
295 usable until about year 3700.
296 Updating from previously releases is fully supported: old-format spool
297 files are still usable, and the utilities support both formats. New
298 message will use the new format. The one hints-DB file type which uses
299 message-IDs (the transport wait- DB) will be discarded if an old-format ID
300 is seen; new ones will be built with only new-format IDs.
301 Optionally, a utility can be used to convert spool files from old to new,
302 but this is only an efficiency measure not a requirement for operation
303 Downgrading from new to old requires running a provided utility, having
304 first stopped all operations. This will convert any spool files from new
305 back to old (losing time-precision and PID information) and remove any
306 wait- hints databases.
308 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
309 we treated them as item separators when parsing for a list item, but they
310 need to be protected by the doublequotes. While there, add handling for
313 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
314 Found and fixed by Jasen Betts. No testcase for this as my usual text
315 editor insists on emitting only valid UTF-8.
317 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
318 a null-indirection SIGSEGV for the receive process.
320 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
321 -bh use. Internal checking would spot this and take a panic.
323 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
324 In 4.96 this would expand to empty.
326 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
327 certificate. Find and fix by Andreas Metzler.
329 JH/36 Add ARC info to DMARC hostory records.
331 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
332 or fakedefer. Previously the sender could discover that the message
333 had in fact been accepted.
335 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
336 sequences. Previously the input was not noted as being tainted; notably
337 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
338 bad coding of authenticators.
340 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
341 and ${tr...}. Found and diagnosed by Heiko Schlichting.
343 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
344 could be triggered by externally-supplied input. Found by Trend Micro.
347 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
348 be triggered by externally-controlled input. Found by Trend Micro.
351 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
352 be triggered by externally-controlled input. Found by Trend Micro.
355 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
356 Make the rewrite never match and keep the logging. Trust the
357 admin to be using verify=header-syntax (to actually reject the message).
359 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
362 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
368 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
369 after reception to before a subsequent reception. This should
370 mean slightly faster delivery, and also confirmation of reception
373 JH/02 Move from using the pcre library to pcre2. The former is no longer
374 being developed or supported (by the original developer).
376 JH/03 Constification work in the filters module required a major version
377 bump for the local-scan API. Specifically, the "headers_charset"
378 global which is visible via the API is now const and may therefore
379 not be modified by local-scan code.
381 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
382 sendfile() didi not account for the way the ClamAV driver code called it.
384 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
385 time check was being done for every character; replace that with one
388 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
389 sent was prefixed with a length byte.
391 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
392 RFC 5321. Previously Dovecot (at least) would log errors during
395 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
396 in 4.61 without a reason logged. Binaries will be bigger, which might
397 matter on diskspace-constrained systems, but debug is easier.
399 JH/09 Fix macro-definition during "-be" expansion testing. The move to
400 write-protected store for macros had not accounted for these runtime
401 additions; fix by removing this protection for "-be" mode.
403 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
404 handing out large-numbered file descriptors, violating the usual Unix
405 assumption (and required by Posix) that the lowest possible number will be
406 allocated by the kernel when a new one is needed. In the daemon, and any
407 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
408 are not useable for FD_SET() [and hence select()] and overwrite the stack.
409 Assorted crashes happen.
411 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
412 main-section options or in a connect ACL, the value from the first ever
413 connection was never replaced for subsequent connections. Found by
416 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
417 though only once PCRE2 was introduced: the memory accounting used under
418 debug offset allocations by an int, giving a hard trap in early startup.
419 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
421 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
422 with underbars is given. The write-protection of configuration introduced
423 in 4.95 trapped when normalisation was applied to an option not needing
426 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
428 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
429 failing to close the certificates directory, every hour or any time it
432 JH/16 Debugging initiated by an ACL control now continues through into routing
433 and transport processes. Previously debugging stopped any time Exim
434 re-execs, or for processing a queued message.
436 JH/17 The "expand" debug selector now gives more detail, specifically on the
437 result of expansion operators and items.
439 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
440 bad comparison between the option value and the name of the file to
441 be included was done, and a mismatch was wrongly identified.
442 4.88 to 4.95 are affected.
444 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
446 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
447 supplied as the db file-pair basename. Previously, if a directory
448 path was given, for example via the autoreply "once" option, the DB
449 file.pag and file.dir files would be created in that directory's
452 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
453 "taint" log_selector. These were previously deprecated.
455 JH/22 Fix static address-list lookups to properly return the matched item.
456 Previously only the domain part was returned.
458 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
459 the call into OpenSSL to send a TLS Close was being repeated; this
460 resulted in the library waiting for the peer's Close. If that was never
461 sent we waited forever. Fix by tracking send calls.
463 JH/24 The ${run} expansion item now expands its command string elements after
464 splitting. Previously it was before; the new ordering makes handling
465 zero-length arguments simpler. The old ordering can be obtained by
466 appending a new option "preexpand", after a comma, to the "run".
468 JH/25 Taint-check exec arguments for transport-initiated external processes.
469 Previously, tainted values could be used. This affects "pipe", "lmtp" and
470 "queryprogram" transport, transport-filter, and ETRN commands.
471 The ${run} expansion is also affected: in "preexpand" mode no part of
472 the command line may be tainted, in default mode the executable name
475 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
476 the the facility was not passed across execs, and only the first message
477 passed over a connection could use BDAT; any further ones using DATA.
479 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
480 uses $sending_ip_address and an interface is specified.
481 Previously any use of the local address in the EHLO name disabled
482 PIPECONNECT, the common case being to use the rDNS of it.
484 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
485 resumption. Previously verify failed because no certificate status is
486 passed on the wire for the restarted session. Fix by using the recorded
487 ocsp status of the stored session for the new connection.
489 JH/29 TLS resumption: the key for session lookup in the client now includes
490 more info that a server could potentially use in configuring a TLS
491 session, avoiding oferring mismatching sessions to such a server.
492 Previously only the server IP was used.
494 JH/30 Fix string_copyn() for limit greater than actual string length.
495 Previously the copied amount was the limit, which could result in a
496 overlapping memcpy for newly allocated destination soon after a
497 source string shorter than the limit. Found/investigated by KM.
499 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
500 close; it may be needed for a subsequent connection. This caused a
501 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
503 JH/32 Fix CHUNKING for a second message on a connection when the first was
504 rejected. Previously we did not reset the chunking-offered state, and
505 erroneously rejected the BDAT command. Investigation help from
508 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
509 an empty address. Previously the expansion returned an error.
511 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
512 proxy. Previously these were misparsed, leading to paniclog entries.
518 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
519 related applications. Previously an "H" was used where available info
520 says that "M" should be, so change to match.
522 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
523 as arguments, so an implementation trying to copy these into a local
524 buffer was taking a taint-enforcement trap. Fix by using dynamically
525 created buffers. Similar fix for radius expansion condition.
527 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
528 reasonable, eg. to count headers. Fix by using dynamically created
529 buffers rather than a local. Do similar fixes for ACL actions "dcc",
530 "log_reject_target", "malware" and "spam"; the arguments are expanded
531 so could be handling tainted values.
533 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
534 broken the (no-op) support for this sendmail command. Restore it
535 to doing nothing, silently, and returning good status.
537 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
538 record path was given (or the default used) without a leading directory
539 path, an error occurred on trying to open it. Use the transport's working
542 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
543 transport. Previously it was the name on the DNS A-record; use instead
544 the head of the CNAME chain leading there (if there is one). This seems
545 to align better with RFC 6125.
547 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
548 smtp_accept_max_per_host allocated resources which were not released
549 when the limit was exceeded. This eventually crashed the daemon. Fix
550 by adding a release action in that path.
552 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
553 expanded; previously using tainted values was rejected. Fix by using
554 dynamically-created buffers.
556 JH/09 Relax restrictions on ACL verify condition needing access to message
557 headers. Previously they were only permitted in data and non-smtp ACLs;
558 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
559 not_blind, header_sender and header_names_ascii verification.
561 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
562 Previously a macro used one argument twice; when called with the
563 argument as an expression having side-effects, incorrect operation
564 resulted. Use an inlineable function.
566 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
567 held open for a verify callout. Previously this wan not accounted for
568 and a corrupt onward SMTP conversation resulted.
570 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
571 Previously they were embedded naively in the constructed address; when
572 needed, strip the quoting and quote the entire local_part.
573 Also make the inbound_srs expansion condition handle quoting.
575 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
576 excluded, not matching the documentation.
578 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
579 was given for the sqlite_dbfile a trap resulted.
581 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
582 "name" argument resulted in a trap. There is no reason to disallow such;
583 this was a coding error.
585 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
586 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
587 spent suspended, ignoring the POSIX definition. Previously we assumed
588 it did and a constant offset from real time could be used as a correction.
589 Change to using the same clock source for the start-of-message and the
590 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
591 exists, just to get a clock slightly more aligned to reality.
593 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
594 RFC says it is optional some validators care. The missing char was not
595 intended but triggered by a line-wrap alignment. Discovery and fix by
596 Guillaume Outters, hacked on by JH.
598 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
599 name being quoted was tainted a trap would be taken. Fix by using
600 dynamically created buffers. The routine could have been called by a
601 rewrite with the "h" flag, by using the "-F" command-line option, or
602 by using a "name=" option on a control=submission ACL modifier.
604 JH/19 SPF: change the Authentication-Results expansion component to give
605 smtp.helo when the sender domain is empty. Previously it gave
608 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
609 not in 127.0.0.0/8 to help in spotting list domains taken over by a
610 domain-parking registrar.
612 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
613 Previously when a whitespace character was specified it was not inserted
614 after removing the newline.
616 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
617 the domain part of the recipient address. This overrides any tls_sni
618 option set, which was previously used.
620 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
623 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
624 is_tainted() had an off-by-one error in the overenthusiastic direction.
625 Find and fix by Gavan. Although NetBSD is not a supported platform for
626 4.94 this bug could affect other platforms.
628 PP/01 Fix default prime selection to be consistent.
629 One path used ike23 still, instead of exim.dev.20160529.3; now both
630 execution flows will use the same DH primes (currently
631 exim.dev.20160529.3).
633 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
634 option in smtp transport, to match the documentation. Previously
635 verification was not being done in some cases where it should have been.
637 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
638 than one server was defined and depending on the platform memory layout
639 details, an internal consistency trap could be hit while walking the list
642 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
643 files. The value is used by the authresults expansion item. Previously
644 if this was used in a router or transport, a crash could result.
646 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
647 have a sufficient incidence of EINTR returns from select that an
648 interaction with other operations done by the main daemon loop exposed
649 a bug in the error-handling. This was benign apart from the log
652 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
653 for consistency with delivered "=>" and failed "**" lines. While we're
654 there, handle PRX and TFO.
656 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
657 applied. This resulted, if any header-line rewrite rules were configured,
658 in a panic-log triggerable by sending a message with a long address in
659 a header. Fix by increasing the arbitrary limit to larger than a single
660 (dewrapped) 5322 header line maximum size.
662 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
663 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
664 RFC 6648 which deprecates X- options in protocols as a general practice.
665 Changeover between the implementations is handled by the mechanisms
668 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
669 a remote which uses callout/hold. Previously the local was queued.
671 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
674 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
675 log line. The asterisk is now consistently appended to the remote IP
676 (and port, if given), and will also be provided on defer and fail log
677 lines. Previously it could be placed on the local IP if that was being
678 logged, and was only provided on delivery lines.
680 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
682 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
683 authenticator client_send option. Previously the next char, after a pair
684 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
685 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
686 documentation. There is still no way to get a leading ^ immediately
687 after a NUL (ie. for the password of a PLAIN method authenticator.
689 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
690 files. For bad sizes read, delete the record and whine to paniclog.
692 JH/38 When logging an AUTH failure, as server, do not include sensitive
693 information. Previously, the credentials would be included if given
694 as part of the AUTH command line and an ACL denied authentication.
696 JH/39 Bug 2691: fix $local_part_data. When the matching list element
697 referred to a file, bad data was returned. This likely also affected
700 JH/40 The gsasl authenticator now supports caching of the salted password
701 generated by the client-side implementation. This required the addition
702 of a new variable: $auth4.
704 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
705 left undeleted; the attempt to re-create it then failed - resulting in
706 the usual "SIGHUP tp have daemon reload configuration" to not work.
707 This affected any platform not supporting "abstract" Unix-domain
708 sockets (i.e. not Linux).
710 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
711 recipients" response to RCPT in a later response, with a 250. The
712 previous coding assumed this would not happen, and under PIPELINING
713 would result in both lost and duplicate recipients for a message.
715 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
716 Previously the weighting was incorrectly applied. Similar fix for socks
717 proxies. Found and fixed by Heiko Schlichting.
719 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
720 not handle sub-lists included using the +namedlist syntax. While
721 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
722 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
725 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
726 Previously the permanent pool was used, so the store could not be freed.
727 This meant a connection with many messages would use continually-growing
730 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
731 per-pool so as not to waste too much space. Previously a constant size
732 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
733 DOS attacks harder. The cost is wasted memory use in the larger blocks.
735 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
736 same space repeatedly, and a smaller process footprint.
738 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
739 for GnuTLS. Previously the name started "%s" which, while not a bug,
740 looked as if if might be one.
742 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
743 "continued connection") make the $proxy_* variables available. Previously
744 the information was not passed across the exec() call for subsequent
745 transport executions. This also mean that the log lines for the
746 messages can show the proxy information.
748 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
749 receive time. With modern systems the difference is significant.
750 The historical behaviour can be restored by disabling (a new) log_selector
751 "queue_time_exclusive".
753 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
754 content) were specifically tested for. Now, also cover expansions
755 resulting in ACL names and inline ACL content.
757 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
758 making it unusable in complex expressions.
760 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
761 fail with a taint-check on the spoolfile name, and leave the message
764 HS/01 Enforce absolute PID file path name.
766 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
768 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
770 PP/02 Bug 2643: Correct TLS DH constants.
771 A missing NUL termination in our code-generation tool had led to some
772 incorrect Diffie-Hellman constants in the Exim source.
773 Reported by kylon94, code-gen tool fix by Simon Arlott.
775 PP/03 Impose security length checks on various command-line options.
776 Fixes CVE-2020-SPRSS reported by Qualys.
778 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
779 better. Reported by Qualys.
781 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
782 providing a particularly obnoxious sender full name.
785 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
787 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
790 PP/08 Change default for recipients_max from unlimited to 50,000.
792 PP/09 Fix security issue with too many recipients on a message (to remove a
793 known security problem if someone does set recipients_max to unlimited,
794 or if local additions add to the recipient list).
795 Fixes CVE-2020-RCPTL reported by Qualys.
797 PP/10 Fix security issue in SMTP verb option parsing
798 Fixes CVE-2020-EXOPT reported by Qualys.
800 PP/11 Fix security issue in BDAT state confusion.
801 Ensure we reset known-good where we know we need to not be reading BDAT
802 data, as a general case fix, and move the places where we switch to BDAT
803 mode until after various protocol state checks.
804 Fixes CVE-2020-BDATA reported by Qualys.
806 HS/03 Die on "/../" in msglog file names
808 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
809 the Exim runtime user are allowed to create files.
811 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
814 QS/03 When reading the output from interpreted forward files we do not
815 pass the pipe between the parent and the interpreting process to
816 executed child processes (if any).
818 QS/04 Always die if requested from internal logging, even is logging is
821 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
822 the API; compilation noo longer completes with DMARC support included.
823 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
824 been reported on other platforms.
826 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
828 JH/56 Make the majority of info read from config files readonly, for defence-in-
829 depth against exploits. Suggestion by Qualys.
830 Not supported on Solaris 10.
832 JH/57 Fix control=fakreject for a custom message containing tainted data.
833 Previously this resulted in a log complaint, due to a re-expansion present
834 since fakereject was originally introduced.
836 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
837 modifier was given, a loop resulted.
839 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
840 pipelined SMTP command followed the BDAT LAST then it would be
841 incorrectly treated as part of the message body, causing a verification
844 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
845 require looser validation rules than those for 821-level addresses,
846 which only permit IP addresses.
852 JH/01 Avoid costly startup code when not strictly needed. This reduces time
853 for some exim process initialisations. It does mean that the logging
854 of TLS configuration problems is only done for the daemon startup.
856 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
858 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
859 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
862 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
863 DKIM signing is being done. Previously a transport_filter would always
864 disable CHUNKING, falling back to traditional DATA.
866 JH/05 Regard command-line recipients as tainted.
868 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
870 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
871 PAM library frees one of the arguments given to it, despite the
872 documentation. Therefore a plain malloc must be used.
874 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
875 on-stack buffers were used, resulting in a taint trap when DSN information
876 copied from a received message was written into the buffer.
878 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
879 the ordering of its ARC headers. This caused a crash.
881 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
882 a new record was being constructed with information from the peer, a trap
885 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
886 installation would get error messages from DMARC verify, when it hit the
887 nonexistent file indicated by the default. Distros wanting DMARC enabled
888 should both provide the file and set the option.
889 Also enforce no DMARC verification for command-line sourced messages.
891 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
892 could, depending on the platform, hang at the STARTTLS response.
894 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
895 message on a connection. Previously if one message had ARC headers and
896 the following one did not, a crash could result when adding an
897 Authentication-Results: header.
899 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
900 Exim main code and Exim-related utities. The introduction of taint
901 tracking also did many adjustments to string handling. Since then, eximon
902 frequently terminated with an assert failure.
904 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
905 check for 452 responses. This slightly helps the inefficieny of doing
906 a large alias-expansion into a recipient-limited target. The max_rcpt
907 transport option still applies (and at the current default, will override
908 the new feature). The check is done for either cause of synch, and forces
909 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
910 connection. The new facility is not tunable at this time.
912 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
913 library live data was being used, so the results became garbage. Make
914 copies while it is still usable.
916 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
917 on delivery deferred (==) and failed (**) lines (if a delivery was
918 attemtped). Previously it was only on completion (=>) lines.
920 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
921 for the expansion of the server_scram_iter and server_scram_salt options.
923 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
924 are now specifically given a NO_DATA response without hitting the system
925 resolver. The library goes on to do the now-standard TXT lookup.
926 Use of dnsdb lookups is not affected.
928 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
929 only retrieve the errormessage once. Previously two calls to dlerror()
930 were used, and the second one (for mainlog/paniclog) retrieved null
933 JH/20 Taint checking: disallow use of tainted data for
934 - the appendfile transport file and directory options
935 - the pipe transport command
936 - the autoreply transport file, log and once options
937 - file names used by the redirect router (including filter files)
939 - paths used by single-key lookups
940 Previously this was permitted.
942 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
943 adjusted the size of a major service buffer; this failed because the
944 buffer was in use at the time. Change to a compile-time increase in the
945 buffer size, when this authenticator is compiled into exim.
947 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
948 previous fast-mode was untenable in the face of glibs using mmap to
949 support larger malloc requests.
951 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
952 New values supported, if defined on system where compiled:
953 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
954 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
956 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
957 running a limited number of proceses in parallel, a benefit is gained. The
958 amount varies with the platform hardware and load. The use of the option
959 queue_run_in_order means we cannot do this, as ordering becomes
962 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
963 had introduced a string-copy (for ensuring NUL-termination) which was not
964 appropriate for that case, which can include embedded NUL bytes in the
965 block of data. Investigation showed the copy to actually be needless, the
966 data being length-specified.
968 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
969 done during a receiving connection, and both used TLS, global info was
970 used rather than per-connection info for tracking the state of data
971 queued for transmission. This could result in a connection hang.
973 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
974 Previously, when delivering serveral messages down a single connection
975 only the first would provide a SIZE. This was due to the size information
976 not being properly tracked.
978 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
979 TAI (at 37 seconds currently), pretend to be in UTC for time-related
980 expansion and logging. Previously, spurious values such as a future
981 minute could be seen.
983 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
984 it could crash from a null-deref. This could also affect the
985 ${addresses: } operator and ${readsock } item.
987 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
988 message following a mime one, the variable was not reset.
990 JH/30 When an pipelined-connect fails at the first response, assume incorrect
991 cached capability (perhaps the peer reneged?) and immediately retry in
994 JH/31 Fix spurious detection of timeout while writing to transport filter.
996 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
997 an attempt to copy the string was made before checking it.
999 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
1000 taint of the lookup key was maintained; we now regard the presence in the
1001 filesystem as sufficient validation.
1003 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
1004 argument is supplied.
1006 JH/35 The dsearch lookup now requires that the directory is an absolute path.
1007 Previously this was not checked, and nonempty relative paths made an
1008 access under Exim's current working directory.
1010 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
1011 Previously no event was raised.
1013 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
1014 parameter supplied by the sender MAIL FROM command. Previously it was
1015 ignored, and only the check_spool_space option value for the required
1018 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
1019 the size of the signing public-key. Previously it was instead giving
1020 the size of the signature hash.
1022 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
1023 the default. See the (new) dkim_verify_min_keysizes option.
1025 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
1026 and an ACL use a lookup for checking either the local_part or domain,
1027 stale data could be accessed. Ensure that variable references are
1028 dropped between messages.
1030 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
1031 by the client was not checked as pointing within response data before
1032 being used. A malicious client could thus cause an out-of-bounds read and
1033 possibly gain authentication. Fix by adding the check.
1035 JH/42 Internationalisation: change the default for downconversion in the smtp
1036 transport to be "if needed". Previously it was "as previously set" for
1037 the message, which usually meant "if needed" for message-submission but
1038 "no" for everything else. However, MTAs have been seen using SMTPUTF8
1039 even when the envelope addresses did not need it, resulting in forwarding
1040 failures to non-supporting MTAs. A downconvert in such cases will be
1041 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
1042 transport. The change does mean that addresses needing conversion will
1043 be converted when previously a delivery failure would occur.
1045 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1046 response was received it would be used unchecked in a fail-DSN, violating
1047 standards on line-length limits. Truncate if needed.
1049 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1050 doesn't work. (Additionally add a new main config option to configure the
1057 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1058 side, to decode a TLS 1.3 packet capture.
1060 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1061 Previously the default library behaviour applied, sending two, each in
1062 its own TCP segment.
1064 JH/03 Debug output for ACL now gives the config file name and line number for
1067 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1069 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1071 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1072 buffer overrun for (non-chunking) other transports.
1074 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1075 TLS1.3, means that a server rejecting a client certificate is not visible
1076 to the client until the first read of encrypted data (typically the
1077 response to EHLO). Add detection for that case and treat it as a failed
1078 TLS connection attempt, so that the normal retry-in-clear can work (if
1079 suitably configured).
1081 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1082 and/or domain. Found and fixed by Jason Betts.
1084 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1085 configuration). If a CNAME target was not a wellformed name pattern, a
1088 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1089 the OS reports them interleaved with other addresses.
1091 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1092 used both for input and for a verify callout, both encrypted, SMTP
1093 responses being sent by the server could be lost. This resulted in
1094 dropped connections and sometimes bounces generated by a peer sending
1097 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1098 string. Previously it was possible to cause undefined behaviour in a
1099 library routine (usually a crash). Found by "zerons".
1101 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1104 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1105 API was removed, so update to use the newer ones.
1107 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1108 any timeout set, is taking a long time. Previously we would hang on to a
1109 rotated logfile "forever" if the input was arriving with long gaps
1110 (a previous attempt to fix addressed lack, for a long time, of initial
1113 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1114 shared (NFS) environment. The length of the tempfile name is now
1115 4 + 16 ("hdr.$message_exim_id") which might break on file
1116 systems which restrict the file name length to lower values.
1117 (It was "hdr.$pid".)
1119 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1120 shared (NFS) environment.
1122 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1123 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1126 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1127 on some platforms for bit 31.
1129 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1130 to changes apparently associated with TLS1.3 handling some of the APIs
1131 previously used were either nonfunctional or inappropriate. Strings
1132 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1133 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1134 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1135 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1136 and the use of specific cipher names in the encrypted= ACL condition.
1138 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1140 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1141 verification result was not updated unless hosts_require_ocsp applied.
1143 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1144 queue_list_requires_admin set to false, non-admin users were denied the
1147 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1148 directory-of-certs mode. Previously they were advertised despite the
1151 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1152 A single TCP connection by a client will now hold a TLS connection open
1153 for multiple message deliveries, by default. Previously the default was to
1156 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1157 default. If built with the facility, DANE will be used. The facility
1158 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1160 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1161 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1162 must be defined and you must still, unless you define DISABLE_TLS, manage
1163 the the include-dir and library-file requirements that go with that
1164 choice. Non-TLS builds are still supported.
1166 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1169 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1170 default. If the platform supports and has the facility enabled, it will
1171 be requested on all coneections.
1173 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1174 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1176 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1178 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1179 success-DSN messages. Previously the From: header was always the default
1180 one for these; the option was ignored.
1182 JH/28 Fix the timeout on smtp response to apply to the whole response.
1183 Previously it was reset for every read, so a teergrubing peer sending
1184 single bytes within the time limit could extend the connection for a
1185 long time. Credit to Qualsys Security Advisory Team for the discovery.
1187 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1188 delivery address, which leaked information of the results of local
1189 forwarding. Change to the original envelope recipient address, per
1192 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1193 requested. Previously not bounce was generated and a log entry of
1194 error ignored was made.
1196 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1198 JH/32 Introduce a general tainting mechanism for values read from the input
1199 channel, and values derived from them. Refuse to expand any tainted
1200 values, to catch one form of exploit.
1202 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1203 was unused and the unexpanded text used for the test. Found and
1204 fixed by Ruben Jenster.
1206 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1207 an attempt to use a TLS library read routine dereffed a nul pointer,
1210 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1211 them in our smtp response.
1213 JH/36 Have the general router option retry_use_local_part default to true when
1214 any of the restrictive preconditions are set (to anything). Previously it
1215 was only for check_local user. The change removes one item of manual
1216 configuration which is required for proper retries when a remote router
1217 handles a subset of addresses for a domain.
1219 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1220 link count into consideration.
1222 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1223 caused the extension of big_buffer, the following lines were ignored.
1225 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1226 accordance with RFC 2308. Previously there was no expiry, so a longlived
1227 receive process (eg. due to ACL delays) versus a short SOA value could
1230 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1232 JH/39 Promote DMARC support to mainline.
1234 JH/40 Bug 2452: Add a References: header to DSNs.
1236 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1237 parameters. The relevant library call is documented as "Deprecated: This
1238 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1239 3.6.0, DH parameters are negotiated following RFC7919."
1241 HS/06 Change the default of dnssec_request_domains to "*"
1243 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1244 carried on and emitted a BDAT command, even when PIPELINING was not
1247 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1248 buffer was used for the filename, resulting in a trap when tainted
1249 arguments (eg. $domain) were used.
1251 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1252 recommended to avoid a possible server-load attack. The feature can be
1253 re-enabled via the openssl_options main cofiguration option.
1255 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1256 for version 4.90 - adding a "more data" boolean to the arguments.
1257 Bumped the ABI version number also, this having been missed previously;
1258 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1259 respect of usage of smtp_printf() by either local_scan code or libraries
1260 accessed via the ${dlfunc } expansion item. Both will need coding
1261 adjustment for any calls to smtp_printf() to match the new function
1262 signature; a FALSE value for the new argument is always safe.
1264 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1265 the file-offset (which the Linux syscall does, and exim expects); this
1266 resulted in an indefinite loop.
1268 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1269 to do ARC verification. The Authentication-Results: header line added
1270 by the configuration then had no ARC item.
1272 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1273 an incorrect mode was used for reading the result, resulting in it being
1280 JH/01 Remove code calling the customisable local_scan function, unless a new
1281 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1283 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1284 non-signal-safe functions being used.
1286 JH/03 Bug 2269: When presented with a received message having a stupidly large
1287 number of DKIM-Signature headers, disable DKIM verification to avoid
1288 a resource-consumption attack. The limit is set at twenty.
1290 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1291 report of oldest_pass in ${authres } in consequence, and separate out
1292 some descriptions of reasons for verification fail.
1294 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1295 files in the spool were present and unlocked. A queue-runner could spot
1296 them, resulting in a duplicate delivery. Fix that by doing the unlock
1297 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1298 add more error-checking on spoolfile handling while that code is being
1301 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1302 No known attacks, no CVE, this is defensive hardening.
1304 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1305 a queue-runner could start a delivery while other operations were ongoing.
1306 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1307 Found and investigated by Tim Stewart. Fix by using the open message data
1308 file handle rather than opening another, and not locally closing it (which
1309 releases a lock) for that case, while creating the temporary .eml format
1310 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1312 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1313 $sender_verify_failure/$recipient_verify_failure to "random".
1315 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1318 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1319 Previously this would segfault.
1321 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1324 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1325 like zero, since the resolver should be doing this for us, But we need one
1326 as a CNAME but no MX presence gets the CNAME returned; we need to check
1327 that doesn't point to an MX to declare it "no MX returned" rather than
1328 "error, loop". A new main option is added so the older capability of
1329 following some limited number of chain links is maintained.
1331 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1333 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1334 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1335 now-deprecated earlier definitions used only specified the range up to TLS
1336 1.2 (in the older-version library docs).
1338 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1340 JH/15 Rework TLS client-side context management. Stop using a global, and
1341 explicitly pass a context around. This enables future use of TLS for
1342 connections to service-daemons (eg. malware scanning) while a client smtp
1343 connection is using TLS; with cutthrough connections this is quite likely.
1345 JH/16 Fix ARC verification to do AS checks in reverse order.
1347 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1349 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1350 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1351 Previously the "utf8" would be re-prepended for every additional message.
1353 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1354 Previously thery were accepted, resulting in issues when attempting to
1355 forward messages to a non-supporting MTA.
1357 PP/02 Let -n work with printing macros too, not just options.
1359 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1360 one parent address was copied, and bogus data was used at delivery-logging
1361 time. Either a crash (after delivery) or bogus log data could result.
1362 Discovery and analysis by Tim Stewart.
1364 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1365 Previously if the string ended mid-character, we did not insert the
1366 promised '?' replacement.
1368 PP/04 Documentation: current string operators work on bytes, not codepoints.
1370 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1371 should pack well giving a smaller memory footprint so better caching and
1372 therefore performance. Group the declarations where this can't be done so
1373 that the byte-sized flag variables are not interspersed among pointer
1374 variables, giving a better chance of good packing by the compiler.
1376 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1377 non-null, to avoid issues with sites running BATV. Previously reports were
1378 sent with an empty envelope sender so looked like bounces.
1380 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1381 The ignore_error flag wasn't being returned from the filter subprocess so
1382 was not set for later routers. Investigation and fix by Matthias Kurz.
1384 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1385 and a msg:complete for the whole, when a message is manually removed using
1386 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1388 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1389 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1390 hope that is portable enough.
1392 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1393 requiring a known-CA anchor certificate; make it now rely entirely on the
1394 TLSA as an anchor. Checking the name on the leaf cert against the name
1395 on the A-record for the host is still done for TA (but not for EE mode).
1397 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1398 included in delivery lines for non-proxied connections, when compiled with
1399 SUPPORT_SOCKS and running with proxy logging enabled.
1401 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1402 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1403 move the existing event to fire before the normal logging of message
1404 failure so that custom logging is bracketed by normal logging.
1406 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1407 msg:fail:internal event. Developement by Matthias Kurz.
1409 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1410 far too small for todays use of crypto signatures stored there. Go all
1411 the way to the max DNS message size of 64kB, even though this might be
1412 overmuch for IOT constrained device use.
1414 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1415 copy a string over itself. The library routine is documented as not
1416 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1418 JH/32 For main options check_spool_space and check_inode_space, where the
1419 platform supports 64b integers, support more than the previous 2^31 kB
1420 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1421 the previous G, M, k.
1423 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1424 $authenticated_fail_id variable on authentication failure. Previously
1427 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1428 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1429 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1430 GNUTLS_SEC_PARAM_MEDIUM.
1432 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1433 side. Previously we would continue as if no SNI had been received.
1435 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1436 "<" character, which should be a list-separator specification, we walked
1437 off past the nul-terimation.
1439 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1440 causes) even when the retry time is not yet met. Previously they were
1441 not, meaning that when (say) an account was over-quota and temp-rejecting,
1442 and multiple senders' messages were queued, only one sender would get
1443 notified on each configured delay_warning cycle.
1445 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1447 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1448 cipher-suites, an error can be left on the stack even for a succeeding
1449 accept; this results in impossible error messages when a later operation
1452 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1453 return error codes indicating retry. Under TLS1.3 this becomes required.
1455 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1456 it only wrote the new authenticators, resulting in a lack of tracking of
1457 peer changes of ESMTP extensions until the next cache flush.
1459 JH/41 Fix the loop reading a message header line to check for integer overflow,
1460 and more-often against header_maxsize. Previously a crafted message could
1461 induce a crash of the recive process; now the message is cleanly rejected.
1463 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1464 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1470 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1471 When redis_servers is set to a list of > 1 element, and the Redis servers
1472 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1473 case of MOVED into a DEFER case instead, thus moving the query onto the
1474 next server in the list. For a cluster of N elements, all N servers must
1475 be defined in redis_servers.
1477 GF/02 Catch and remove uninitialized value warning in exiqsumm
1478 Check for existence of @ARGV before looking at $ARGV[0]
1480 JH/01 Replace the store_release() internal interface with store_newblock(),
1481 which internalises the check required to safely use the old one, plus
1482 the allocate and data copy operations duplicated in both (!) of the
1483 extant use locations.
1485 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1486 modifier. This matches the restriction on the commandline.
1488 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1489 Previously only the last row was returned.
1491 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1492 we assumed that tags in the header were well-formed, and parsed the
1493 element content after inspecting only the first char of the tag.
1494 Assumptions at that stage could crash the receive process on malformed
1497 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1498 While running the DKIM ACL we operate on the Permanent memory pool so that
1499 variables created with "set" persist to the DATA ACL. Also (at any time)
1500 DNS lookups that fail create cache records using the Permanent pool. But
1501 expansions release any allocations made on the current pool - so a dnsdb
1502 lookup expansion done in the DKIM ACL releases the memory used for the
1503 DNS negative-cache, and bad things result. Solution is to switch to the
1504 Main pool for expansions.
1505 While we're in that code, add checks on the DNS cache during store_reset,
1506 active in the testsuite.
1507 Problem spotted, and debugging aided, by Wolfgang Breyha.
1509 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1510 When none of the hosts presented to a transport match an already-open
1511 connection, close it and proceed with the list. Previously we would
1512 queue the message. Spotted by Lena with Yahoo, probably involving
1515 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1516 Previously a spurious "250 OK id=" response was appended to the proper
1519 JH/08 The "support for" informational output now, which built with Content
1520 Scanning support, has a line for the malware scanner interfaces compiled
1521 in. Interface can be individually included or not at build time.
1523 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1524 by the template makefile "src/EDITME". The "STREAM" support for an older
1525 ClamAV interface method is removed.
1527 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1528 rows affected is given instead).
1530 JH/11 The runtime Berkeley DB library version is now additionally output by
1531 "exim -d -bV". Previously only the compile-time version was shown.
1533 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1534 SMTP connection. Previously, when one had more recipients than the
1535 first, an abortive onward connection was made. Move to full support for
1536 multiple onward connections in sequence, handling cutthrough connection
1537 for all multi-message initiating connections.
1539 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1540 routers. Previously, a multi-recipient message would fail to match the
1541 onward-connection opened for the first recipient, and cause its closure.
1543 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1544 a timeout on read on a GnuTLS initiating connection, resulting in the
1545 initiating connection being dropped. This mattered most when the callout
1546 was marked defer_ok. Fix to keep the two timeout-detection methods
1549 JH/15 Relax results from ACL control request to enable cutthrough, in
1550 unsupported situations, from error to silently (except under debug)
1551 ignoring. This covers use with PRDR, frozen messages, queue-only and
1554 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1556 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1557 metadata, resulting in a crash in free().
1559 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1560 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1561 Broken also in d185889f4, with init system revamp.
1563 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1564 Previously we abruptly closed the connection after reading a malware-
1565 found indication; now we go on to read the "scan ok" response line,
1568 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1569 ACL. Previously, a crash would result.
1571 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1572 macro text after a replacement (previously it was only once per line) and
1573 by skipping builtin macros when searching for an uppercase lead character.
1575 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1576 for the build is renamed.
1578 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1579 was allocated for every new TLS startup, meaning one per message. Fix
1580 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1582 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1583 reported the original. Fix to report (as far as possible) the ACL
1584 result replacing the original.
1586 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1587 OpenSSL. Certificate information is loaded for every new TLS startup,
1588 and the resources needed to be freed.
1590 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1592 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1595 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1596 DATA response info to the (existing) per-recipient response info for
1597 the "C=" log element. It can have useful tracking info from the
1598 destination system. Patch from Simon Arlott.
1600 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1601 length value. Previously this would segfault.
1603 HS/02 Support Avast multiline protoocol, this allows passing flags to
1604 newer versions of the scanner.
1606 JH/28 Ensure that variables possibly set during message acceptance are marked
1607 dead before release of memory in the daemon loop. This stops complaints
1608 about them when the debug_store option is enabled. Discovered specifically
1609 for sender_rate_period, but applies to a whole set of variables.
1610 Do the same for the queue-runner and queue-list loops, for variables set
1611 from spool message files. Do the same for the SMTP per-message loop, for
1612 certain variables indirectly set in ACL operations.
1614 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1615 as a multi-recipient message from a mailinglist manager). The coding had
1616 an arbitrary cutoff number of characters while checking for more input;
1617 enforced by writing a NUL into the buffer. This corrupted long / fast
1618 input. The problem was exposed more widely when more pipelineing of SMTP
1619 responses was introduced, and one Exim system was feeding another.
1620 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1621 receiving system, and refused recipients seen by the sending system
1622 (propating to people being dropped from mailing lists).
1623 Discovered and pinpointed by David Carter.
1625 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1626 replaced by the ${authresults } expansion.
1628 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1630 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1631 allows proper process termination in container environments.
1633 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1634 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1636 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1637 and "err_perm", deprecated since 4.83 when the RFC-defined words
1638 "temperror" and "permerror" were introduced.
1640 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1641 transport-filters or DKIM-signing. The restriction was lost in the
1642 consolidation of verify-callout and delivery SMTP handling.
1643 Extend the restriction to also cover ARC-signing.
1645 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1646 in defer=pass mode supply a 450 to the initiator. Previously the message
1649 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1650 tls_require_ciphers is used as before.
1652 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1653 "pass_unscanned". Only tmpfails from the scanner are written to
1654 the paniclog, as they may require admin intervention (permission
1655 denied, license issues). Other scanner errors (like decompression
1656 bombs) do not cause a paniclog entry.
1658 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1659 Previously it was possible to log spurious information in receive log
1662 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1663 triggered odd behaviour from Outlook Express clients.
1665 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1668 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1669 since the IETF WG has not yet settled on that versus the original
1670 "bare" representation.
1672 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1673 Previously the millisecond value corrupted the output.
1674 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1675 corrupted the output.
1681 JH/01 Rework error string handling in TLS interface so that the caller in
1682 more cases is responsible for logging. This permits library-sourced
1683 string to be attached to addresses during delivery, and collapses
1684 pairs of long lines into single ones.
1686 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1687 during configuration. Wildcards are allowed and expanded.
1689 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1690 This permits better logging.
1692 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1693 do not close it down and have the child transport start it up again on
1694 the passed-on TCP connection. Instead, proxy the child (and any
1695 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1696 affected: the continued delivery log lines do not have any DNSSEC, TLS
1697 Certificate or OCSP information. TLS cipher information is still logged.
1699 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1700 identical IP addresses on different listening ports. Will also affect
1703 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1704 add noisy ifdef guards to special-case this sillyness.
1705 Patch from Bernd Kuhls.
1707 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1708 than 255 are no longer allowed.
1710 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1711 Disable the session-cache too, which might reduce our load. Since we
1712 currrectly use a new context for every connection, both as server and
1713 client, there is no benefit for these.
1714 GnuTLS appears to not support tickets server-side by default (we don't
1715 call gnutls_session_ticket_enable_server()) but client side is enabled
1716 by default on recent versions (3.1.3 +) unless the PFS priority string
1719 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1720 <https://reproducible-builds.org/specs/source-date-epoch/>.
1722 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1723 the check for any unsuccessful recipients did not notice the limit, and
1724 erroneously found still-pending ones.
1726 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1727 MSG_MORE. Only in-clear (not on TLS connections).
1729 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1730 transport-filter is involved we can buffer the headers in memory for
1731 creating the signature, and read the spool data file once for the
1732 signature and again for transmission.
1734 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1735 4.77 as the kernel support then wasn't solid, having issues in 64bit
1736 mode. Now, it's been long enough. Add support for FreeBSD also.
1738 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1739 case where the routing stage had gathered several addresses to send to
1740 a host before calling the transport for the first, we previously failed
1741 to close down TLS in the old transport process before passing the TCP
1742 connection to the new process. The new one sent a STARTTLS command
1743 which naturally failed, giving a failed delivery and bloating the retry
1744 database. Investigation and fix prototype from Wolfgang Breyha.
1746 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1747 false-negatives in the check that the sender had not preempted a response
1748 or prompt from Exim (running as a server), due to that code's lack of
1749 awareness of the SMTP input buffering.
1751 PP/04 Add commandline_checks_require_admin option.
1752 Exim drops privileges sanely, various checks such as -be aren't a
1753 security problem, as long as you trust local users with access to their
1754 own account. When invoked by services which pass untrusted data to
1755 Exim, this might be an issue. Set this option in main configuration
1756 AND make fixes to the calling application, such as using `--` to stop
1759 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1760 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1761 responses to those, into a single TLS record each way (this usually means
1762 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1765 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1766 affects you only if you're dancing at the edge of the param size limits.
1767 If you are, and this message makes sense to you, then: raise the
1768 configured limit or use OpenSSL 1.1. Nothing we can do for older
1771 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1772 cmdline element to get the documented default one. Previously it was
1775 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1776 are unsuable for non-message-specific reasons - eg. connection timeout,
1777 banner-time rejection.
1779 JH/16 Fix logging of delivery remote port, when specified by router, under
1782 PP/06 Repair manualroute's ability to take options in any order, even if one
1783 is the name of a transport.
1786 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1788 JH/17 Change the list-building routines interface to use the expanding-string
1789 triplet model, for better allocation and copying behaviour.
1791 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1792 Previously it was constructed the first time a possibly-matching string
1793 was met in the configuration file input during startup; now it is done
1796 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1797 compatible one, to avoid the (poorly documented) possibility of a config
1798 file in the working directory redirecting the DB files, possibly correpting
1799 some existing file. CVE-2017-10140 assigned for BDB.
1801 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1802 cache-hot. Previously, although the result was properly cached, the
1803 initial verify call returned a defer.
1805 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1806 the main verify for receipient in uncached-mode.
1808 JH/22 Retire historical build files to an "unsupported" subdir. These are
1809 defined as "ones for which we have no current evidence of testing".
1811 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1812 if present. Previously it was ignored.
1814 JH/24 Start using specified-initialisers in C structure init coding. This is
1815 a C99 feature (it's 2017, so now considered safe).
1817 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1818 if was a fixed-sized field and bitmask ops via macros; it is now more
1821 PP/07 GitHub PR 56: Apply MariaDB build fix.
1822 Patch provided by Jaroslav Škarvada.
1824 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1825 during Coverity cleanups [4.87 JH/47]
1826 Diagnosis and fix provided by Michael Fischer v. Mollard.
1828 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1829 the right size to place the terminating semicolon on its own folded
1830 line, the header hash was calculated to an incorrect value thanks to
1831 the (relaxed) space the fold became.
1833 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1834 and confused the parent.
1836 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1837 which could crash as a result. This could lead to undeliverable messages.
1839 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1842 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1843 matters on fast-turnover and PID-randomising systems, which were getting
1844 out-of-order delivery.
1846 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1847 a possibly-overlapping copy. The symptom was that "Remote host closed
1848 connection in response to HELO" was logged instead of the actual 4xx
1851 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1852 Previously only that bufferd was discarded, resulting in SYMTP command
1855 JH/32 DKIM: when a message has multiple signatures matching an identity given
1856 in dkim_verify_signers, run the dkim acl once for each. Previously only
1857 one run was done. Bug 2189.
1859 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1860 "panic the current process" to "deliberately defer". The panic log is
1861 still written with the problem list name; the mail and reject logs now
1862 get a temp-reject line for the message that was being handled, saying
1863 something like "domains check lookup or other defer". The SMTP 451
1864 message is still "Temporary local problem".
1866 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1867 A crafted sequence of BDAT commands could result in in-use memory beeing
1868 freed. CVE-2017-16943.
1870 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1871 from SMTP input. Previously it was always done; now only done for DATA
1872 and not BDAT commands. CVE-2017-16944.
1874 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1875 to the message (such as an overlong header line). Previously this was
1876 not done and we did not exit BDAT mode. Followon from the previous item
1877 though a different problem.
1883 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1884 than -2003 did; needs libidn2 in addition to libidn.
1886 JH/02 The path option on a pipe transport is now expanded before use.
1888 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1889 Patch provided by "Björn", documentation fix added too.
1891 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1892 missing a wire-to-host endian conversion.
1894 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1895 close after a BDAT command line could be taken as a following command,
1896 giving a synch failure. Fix by only checking for synch immediately
1897 before acknowledging the chunk.
1899 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1900 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1901 macro. Patches provided by Josh Soref.
1903 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1904 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1905 by IANA as a service extension. However, John Klensin suggests that we
1908 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1909 the dkim code may be unix-mode line endings rather than smtp wire-format
1910 CRLF, so prepend a CR to any bare LF.
1912 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1913 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1915 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1916 the first were themselves being wrongly included in the feed into dkim
1917 processing; with most chunk sizes in use this resulted in an incorrect
1918 body hash calculated value.
1920 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1921 DKIM signature block, for verification. Although advised against by
1922 standards it is specifically not ruled illegal.
1924 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1926 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1927 missing a body hash (the bh= tag).
1929 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1930 It seems that HAProxy sends the Proxy Protocol information in clear and
1931 only then does a TLS startup, so do the same.
1933 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1934 TCP connections (such as for Spamd) unless the daemon successfully set
1935 Fast Open mode on its listening sockets. This fixes breakage seen on
1936 too-old kernels or those not configured for Fast Open, at the cost of
1937 requiring both directions being enabled for TFO, and TFO never being used
1938 by non-daemon-related Exim processes.
1940 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1941 endings, at least on the first header line. Try to canonify any that get
1942 past that check, despite the cost.
1944 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1945 now limited to an arbitrary five deep, while parsing addresses with the
1946 strip_excess_angle_brackets option enabled.
1948 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1949 instead leave the unprompted TLS handshake in socket buffer for the
1950 TLS library to consume.
1952 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1954 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1956 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1957 region in one operation, for speed, and this leaves assigned pointers
1958 dangling. Add checks run only under the testsuite which checks all
1959 variables at a store-reset and panics on a dangling pointer; add code
1960 explicitly nulling out all the variables discovered. Fixes one known
1961 bug: a transport crash, where a dangling pointer for $sending_ip_address
1962 originally assigned in a verify callout, is re-used.
1964 PP/06 Drop '.' from @INC in various Perl scripts.
1966 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1968 PP/08 Reduce a number of compilation warnings under clang; building with
1969 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1970 should be warning-free.
1972 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1974 HS/01 Fix portability problems introduced by PP/08 for platforms where
1975 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1977 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1978 chunk. This allows us to accept broken chunked messages. We need a more
1979 general solution here.
1981 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1982 already-broken messages in the queue.
1984 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1986 JH/19 Fix reference counting bug in routing-generated-address tracking.
1992 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1993 supports it and a size is available (ie. the sending peer gave us one).
1995 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1996 years of being deprecated). The replacements are the ACLs
1997 acl_smtp_mime and acl_not_smtp_mime.
1999 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
2000 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
2001 or even an in-clear connection were permitted. Now, if the host lookup
2002 was dnssec and dane was requested then the host is only used if the
2003 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
2004 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
2005 if one fails this test.
2006 This means that a poorly-configured remote DNS will make it incommunicado;
2007 but it protects against a DNS-interception attack on it.
2009 JH/04 Bug 1810: make continued-use of an open smtp transport connection
2010 non-noisy when a race steals the message being considered.
2012 JH/05 If main configuration option tls_certificate is unset, generate a
2013 self-signed certificate for inbound TLS connections.
2015 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
2016 in rewrites and routers.
2018 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
2019 and logged a warning sing 4.83; now they are a configuration file error.
2021 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
2022 (lacking @domain). Apply the same qualification processing as RCPT.
2024 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
2026 JH/10 Support ${sha256:} applied to a string (as well as the previous
2029 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
2030 a cutthrough deliver is pending, as we always want to make a connection.
2031 This also avoids re-routing the message when later placing the cutthrough
2032 connection after a verify cache hit.
2033 Do not update it with the verify result either.
2035 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
2036 when routing results in more than one destination address.
2038 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
2039 signing (which inhibits the cutthrough capability). Previously only
2040 the presence of an option was tested; now an expansion evaluating as
2041 empty is permissible (obviously it should depend only on data available
2042 when the cutthrough connection is made).
2044 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
2045 the relevant preceding SMTP command did not note the pipelining mode.
2047 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2048 Previously they were not counted.
2050 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2051 as one having no matching records. Previously we deferred the message
2052 that needed the lookup.
2054 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2055 distinguished as "(=".
2057 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2058 for missing MX records. Previously it only worked for missing A records.
2060 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2062 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2063 after the data-go-ahead and data-ack. Patch from Jason Betts.
2065 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2066 even for a "none" policy. Patch from Tony Meyer.
2068 JH/22 Fix continued use of a connection for further deliveries. If a port was
2069 specified by a router, it must also match for the delivery to be
2072 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2073 When one of the recipients of a message was unsuitable for the connection
2074 (has no matching addresses), we lost track of needing to mark it
2075 deferred. As a result mail would be lost.
2077 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2079 JH/25 Decoding ACL controls is now done using a binary search; the source code
2080 takes up less space and should be simpler to maintain. Merge the ACL
2081 condition decode tables also, with similar effect.
2083 JH/26 Fix problem with one_time used on a redirect router which returned the
2084 parent address unchanged. A retry would see the parent address marked as
2085 delivered, so not attempt the (identical) child. As a result mail would
2088 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2089 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2090 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2093 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2094 Main config options check_{log,spool}_{inodes,space} are now
2095 100 inodes, 10MB unless set otherwise in the configuration.
2097 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2098 Previously it only applied to the main-section connection policy
2101 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2103 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2104 by me. Added RFC7919 DH primes as an alternative.
2106 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2107 are not in the system include path.
2109 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2110 GnuTLS, when a session startup failed (eg because the client disconnected)
2111 Exim did stdio operations after fclose. This was exposed by a recent
2112 change which nulled out the file handle after the fclose.
2114 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2115 signed directly by the cert-signing cert, rather than an intermediate
2116 OCSP-signing cert. This is the model used by LetsEncrypt.
2118 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2120 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2121 an incoming connection.
2123 HS/02 Bug 1802: Do not half-close the connection after sending a request
2126 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2127 fallback to "prime256v1".
2129 JH/34 SECURITY: Use proper copy of DATA command in error message.
2130 Could leak key material. Remotely exploitable. CVE-2016-9963.
2136 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2137 and 3.4.4 - once the server is enabled to respond to an OCSP request
2138 it does even when not requested, resulting in a stapling non-aware
2139 client dropping the TLS connection.
2141 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2142 support variable-length bit vectors. No functional change.
2144 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2145 The I= interface field on outgoing lines is now after the H= remote
2146 host field, same as incoming lines. There is a separate
2147 outgoing_interface log selector which allows you to disable the
2150 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2151 If not running log_selector +smtp_connection the mainlog would be held
2152 open indefinitely after a "too many connections" event, including to a
2153 deleted file after a log rotate. Leave the per net connection logging
2154 leaving it open for efficiency as that will be quickly detected by the
2155 check on the next write.
2157 HS/01 Bug 1671: Fix post transport crash.
2158 Processing the wait-<transport> messages could crash the delivery
2159 process if the message IDs didn't exist for some reason. When
2160 using 'split_spool_directory=yes' the construction of the spool
2161 file name failed already, exposing the same netto behaviour.
2163 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2164 mime_regex ACL conditions.
2166 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2167 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2168 to HELO, local diagnostic string.
2170 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2171 log line to debug. Even when configured with a tls authenticator many
2172 client connections are expected to not authenticate in this way, so
2173 an authenticate fail is not an error.
2175 HS/02 Add the Exim version string to the process info. This way exiwhat
2176 gives some more detail about the running daemon.
2178 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2179 matter for fast-change records such as DNSBLs.
2181 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2182 retry record, even if constant. There may be multiple transports with
2183 different interface settings and the retry behaviour needs to be kept
2186 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2188 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2190 JH/10 Bug 840: fix log_defer_output option of pipe transport
2192 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2193 hosts_randomize. This matters a lot when combined with mua_wrapper.
2195 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2196 ${quote_pgsql:<string>} operator.
2198 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2200 JH/14 Change header file name for freeradius-client. Relevant if compiling
2201 with Radius support; from the Gentoo tree and checked under Fedora.
2203 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2205 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2206 option transport_filter, meaning no filtering.
2208 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2210 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2211 defaults to "*" (all hosts). The variable is now available when not built
2212 with TLS, default unset, mainly to enable keeping the testsuite sane.
2213 If a server certificate is not supplied (via tls_certificate) an error is
2214 logged, and clients will find TLS connections fail on startup. Presumably
2215 they will retry in-clear.
2216 Packagers of Exim are strongly encouraged to create a server certificate
2217 at installation time.
2219 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2220 with the $config_file variable.
2222 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2223 in transport context, after the attempt, and per-recipient. The latter type
2224 is per host attempted. The event data is the error message, and the errno
2225 information encodes the lookup type (A vs. MX) used for the (first) host,
2226 and the trailing two digits of the smtp 4xx response.
2228 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2229 to write to mainlog (or rejectlog, paniclog) in the window between file
2230 creation and permissions/ownership being changed. Particularly affects
2231 installations where exicyclog is run as root, rather than exim user;
2232 result is that the running daemon panics and dies.
2234 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2236 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2237 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2238 "pri" and "weight". Note that the previous implicit priority given by the
2239 list order is no longer honoured.
2241 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2242 for DKIM processing.
2244 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2245 by defining SUPPORT_SOCKS.
2247 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2248 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2249 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2250 variables are renamed to proxy_{local,external}_{address,port}.
2252 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2253 for a build by defining SUPPORT_I18N
2255 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2256 of the query string, and make ${quote_redis:} do that quoting.
2258 JH/29 Move Events support from Experimental to mainline, enabled by default
2259 and removable for a build by defining DISABLE_EVENT.
2261 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2263 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2264 cached by the daemon.
2266 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2267 by defining LOOKUP_REDIS. The libhiredis library is required.
2269 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2270 keys are given for lookup.
2272 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2273 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2274 only supported when built with TLS support. The PolarSSL SHA routines
2275 are still used when the TLS library is too old for convenient support.
2277 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2278 openssl_options), for security. OpenSSL forces this from version 1.1.0
2279 server-side so match that on older versions.
2281 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2282 allocation for $value could be released as the expansion processing
2283 concluded, but leaving the global pointer active for it.
2285 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2286 and to use the domains and local_parts ACL conditions.
2288 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2289 incorrectly not doubled on cutthrough transmission, hence seen as a
2290 body-termination at the receiving system - resulting in truncated mails.
2291 Commonly the sender saw a TCP-level error, and retransmitted the message
2292 via the normal store-and-forward channel. This could result in duplicates
2293 received - but deduplicating mailstores were liable to retain only the
2294 initial truncated version.
2296 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2298 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2300 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2301 we're in there, support oversigning also; bug 1309.
2303 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2305 HS/04 Add support for keep_environment and add_environment options.
2307 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2308 either intentional arithmetic overflow during PRNG, or testing config-
2311 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2312 delivery resulted in actual delivery. Cancel cutthrough before DATA
2315 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2316 rejects a recipient: pass the reject to the originator.
2318 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2319 Many were false-positives and ignorable, but it's worth fixing the
2322 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2323 for the new environment-manipulation done at startup. Move the routines
2324 from being local to tls.c to being global via the os.c file.
2326 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2327 an extract embedded as result-arg for a map, the first arg for extract
2328 is unavailable so we cannot tell if this is a numbered or keyed
2329 extraction. Accept either.
2335 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2338 JH/02 The smtp transport option "multi_domain" is now expanded.
2340 JH/03 The smtp transport now requests PRDR by default, if the server offers
2343 JH/04 Certificate name checking on server certificates, when exim is a client,
2344 is now done by default. The transport option tls_verify_cert_hostnames
2345 can be used to disable this per-host. The build option
2346 EXPERIMENTAL_CERTNAMES is withdrawn.
2348 JH/05 The value of the tls_verify_certificates smtp transport and main options
2349 default to the word "system" to access the system default CA bundle.
2350 For GnuTLS, only version 3.0.20 or later.
2352 JH/06 Verification of the server certificate for a TLS connection is now tried
2353 (but not required) by default. The verification status is now logged by
2354 default, for both outbound TLS and client-certificate supplying inbound
2357 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2360 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2361 Status Notification (bounce) messages are now MIME format per RFC 3464.
2362 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2363 under the control of the dsn_advertise_hosts option, and routers may
2364 have a dsn_lasthop option.
2366 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2367 default, modifiable by a malware= option. The list separator for
2368 the options can now be changed in the usual way. Bug 68.
2370 JH/10 The smtp_receive_timeout main option is now expanded before use.
2372 JH/11 The incoming_interface log option now also enables logging of the
2373 local interface on delivery outgoing connections.
2375 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2376 if the interface and destination host and port all match.
2378 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2381 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2382 Patch from Andrew Lewis.
2384 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2385 now supports optional time-restrictions, weighting, and priority
2386 modifiers per server. Patch originally by <rommer@active.by>.
2388 JH/16 The spamd_address main option now supports a mixed list of local
2389 and remote servers. Remote servers can be IPv6 addresses, and
2390 specify a port-range.
2392 JH/17 Bug 68: The spamd_address main option now supports an optional
2393 timeout value per server.
2395 JH/18 Bug 1581: Router and transport options headers_add/remove can
2396 now have the list separator specified.
2398 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2401 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2404 JH/21 Support for the A6 type of dns record is withdrawn.
2406 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2407 rather than the verbs used.
2409 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2410 from 255 to 1024 chars.
2412 JH/24 Verification callouts now attempt to use TLS by default.
2414 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2415 are generic router options now. The defaults didn't change.
2417 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2418 Original patch from Alexander Shikoff, worked over by JH.
2420 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2421 files if ZCAT_COMMAND is not executable.
2423 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2425 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2427 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2428 Normally benign, it bites when the pair was led to by a CNAME;
2429 modern usage is to not canonicalize the domain to a CNAME target
2430 (and we were inconsistent anyway for A-only vs AAAA+A).
2432 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2434 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2435 when evaluating $sender_host_dnssec.
2437 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2438 $sender_helo_dnssec variable.
2440 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2442 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2444 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2446 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2447 documented as working, but never had. Support all but $spam_report.
2449 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2450 added for tls authenticator.
2452 HS/03 Add perl_taintmode main config option
2458 TL/01 When running the test suite, the README says that variables such as
2459 no_msglog_check are global and can be placed anywhere in a specific
2460 test's script, however it was observed that placement needed to be near
2461 the beginning for it to behave that way. Changed the runtest perl
2462 script to read through the entire script once to detect and set these
2463 variables, reset to the beginning of the script, and then run through
2464 the script parsing/test process like normal.
2466 TL/02 The BSD's have an arc4random API. One of the functions to induce
2467 adding randomness was arc4random_stir(), but it has been removed in
2468 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2469 function when detected.
2471 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2472 cause callback expansion.
2474 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2475 syntax errors in an expansion can be treated as a string instead of
2476 logging or causing an error, due to the internal use of bool_lax
2477 instead of bool when processing it.
2479 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2480 server certificates when making smtp deliveries.
2482 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2484 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2486 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2488 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2489 Merged patch from Sebastian Wiedenroth.
2491 JH/05 Fix results-pipe from transport process. Several recipients, combined
2492 with certificate use, exposed issues where response data items split
2493 over buffer boundaries were not parsed properly. This eventually
2494 resulted in duplicates being sent. This issue only became common enough
2495 to notice due to the introduction of connection certificate information,
2496 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2498 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2499 size buffer was used, resulting in syntax errors when an expansion
2502 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2503 version 3.3.6 or later.
2505 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2506 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2507 both become "event_action", the variables become $event_name, $event_data
2508 and $event_defer_errno. There is a new variable $verify_mode, usable in
2509 routers, transports and related events. The tls:cert event is now also
2510 raised for inbound connections, if the main configuration event_action
2513 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2514 early OCSP support, but no stapling (appears to be less than 1.0.0).
2516 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2517 server certificate names available under the smtp transport option
2518 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2521 JH/10 Time-related extraction expansions from certificates now use the main
2522 option "timezone" setting for output formatting, and are consistent
2523 between OpenSSL and GnuTLS compilations. Bug 1541.
2525 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2526 encoded parameter in the incoming message. Bug 1558.
2528 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2529 include certificate info, eximon was claiming there were spoolfile
2532 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2534 JH/14 Log delivery-related information more consistently, using the sequence
2535 "H=<name> [<ip>]" wherever possible.
2537 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2538 are problematic for Debian distribution, omit them from the release
2541 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2543 JH/16 Fix string representation of time values on 64bit time_t architectures.
2546 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2547 output list separator was used.
2552 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2553 checkers that were complaining about end of non-void function with no
2556 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2557 This was a regression introduced in 4.83 by another bugfix.
2559 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2561 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2562 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2568 TF/01 Correctly close the server side of TLS when forking for delivery.
2570 When a message was received over SMTP with TLS, Exim failed to clear up
2571 the incoming connection properly after forking off the child process to
2572 deliver the message. In some situations the subsequent outgoing
2573 delivery connection happened to have the same fd number as the incoming
2574 connection previously had. Exim would try to use TLS and fail, logging
2575 a "Bad file descriptor" error.
2577 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2578 utilities have not been installed.
2580 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2581 temporary space as the ACL may create new global variables.
2583 TL/01 LDAP support uses per connection or global context settings, depending
2584 upon the detected version of the libraries at build time.
2586 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2587 to extract and use the src ip:port in logging and expansions as if it
2588 were a direct connection from the outside internet. PPv2 support was
2589 updated based on HAProxy spec change in May 2014.
2591 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2593 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2594 Properly escape header and check for NULL return.
2596 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2599 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2601 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2602 characters in header names, implemented as a verify condition.
2603 Contributed by Michael Fischer v. Mollard.
2605 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2606 results permerror and temperror. Previous values are deprecated but
2607 still accepted. In a future release, err_perm and err_temp will be
2608 completely removed, which will be a backward incompatibility if the
2609 ACL tests for either of these two old results. Patch contributed by
2610 user bes-internal on the mailing list.
2612 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2614 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2615 selectors, in both main and reject logs.
2617 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2620 JH/07 Add malware type "sock" for talking to simple daemon.
2622 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2624 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2625 routers/transports under cutthrough routing.
2627 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2628 numbers. Touch up "bool" conditional to keep the same definition.
2630 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2632 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2634 JH/12 Expand items in router/transport headers_add or headers_remove lists
2635 individually rather than the list as a whole. Bug 1452.
2637 Required for reasonable handling of multiple headers_ options when
2638 they may be empty; requires that headers_remove items with embedded
2639 colons must have them doubled (or the list-separator changed).
2641 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2642 view the policy declared in the DMARC record. Currently, $dmarc_status
2643 is a combined value of both the record presence and the result of the
2646 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2648 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2649 dnslookup router and the smtp transport (applying to the forward
2652 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2653 of ldap servers used for a specific lookup. Patch provided by Heiko
2656 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2657 New variable $lookup_dnssec_authenticated for observability.
2659 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2660 Patch submitted by Lars Timman.
2662 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2664 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2665 Requires trusted mode and valid format message id, aborts otherwise.
2666 Patch contributed by Heiko Schlichting.
2668 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2669 certextract with support for various fields. Bug 1358.
2671 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2672 is requested by default, modifiable by smtp transport option
2675 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2676 operate on certificate variables to give certificate fingerprints
2677 Also new ${sha256:cert_variable}.
2679 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2681 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2682 Christian Aistleitner.
2684 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2686 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2687 file. Patch from Wolfgang Breyha.
2689 JH/25 Expand the coverage of the delivery $host and $host_address to
2690 client authenticators run in verify callout. Bug 1476.
2692 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2693 align with daemon_smtp_ports. Bug 72.
2695 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2696 support and error reporting did not work properly.
2698 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2699 and is readable. Patch from Andrew Colin Kissa.
2701 TL/14 Enhance documentation of ${run expansion and how it parses the
2702 commandline after expansion, particularly in the case when an
2703 unquoted variable expansion results in an empty value.
2705 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2707 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2708 by renaming away from it. Addresses GH issue 15, reported by
2711 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2713 TL/15 SECURITY: prevent double expansion in math comparison functions
2714 (can expand unsanitized data). Not remotely exploitable.
2721 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2723 PP/02 Make -n do something, by making it not do something.
2724 When combined with -bP, the name of an option is not output.
2726 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2729 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2730 $sender_host_name and config options to manage this, and basic check
2733 PP/05 DSCP support for outbound connections and control modifier for inbound.
2735 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2736 (Only plugin which currently uses this is kerberos4, which nobody should
2737 be using, but we should make it available and other future plugins might
2738 conceivably use it, even though it would break NAT; stuff *should* be
2739 using channel bindings instead).
2741 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2742 name; added for Sendmail compatibility; requires admin caller.
2743 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2744 just ignore it); requires trusted caller.
2745 Also parse but ignore: -Ac -Am -X<logfile>
2748 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2750 TL/02 Add +smtp_confirmation as a default logging option.
2752 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2753 Patch by Magnus Holmgren from 2007-02-20.
2755 TL/04 Bugzilla 1281 - Spec typo.
2756 Bugzilla 1283 - Spec typo.
2757 Bugzilla 1290 - Spec grammar fixes.
2759 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2761 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2763 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2764 mailing list by by Dmitry Isaikin.
2766 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2768 JH/02 Support "G" suffix to numbers in ${if comparisons.
2770 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2772 NM/01 Bugzilla 1197 - Spec typo
2773 Bugzilla 1196 - Spec examples corrections
2775 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2777 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2778 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2781 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2782 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2784 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2785 "acl {{name}{arg}...}", and optional args on acl condition
2788 JH/05 Permit multiple router/transport headers_add/remove lines.
2790 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2792 JH/07 Avoid using a waiting database for a single-message-only transport.
2793 Performance patch from Paul Fisher. Bugzilla 1262.
2795 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2798 JH/09 Add $headers_added variable, with content from use of ACL modifier
2799 add_header (but not yet added to the message). Bugzilla 199.
2801 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2802 Pulled from Bugzilla 817 by Wolfgang Breyha.
2804 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2806 (nb: this is the same fix as in Exim 4.80.1)
2808 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2811 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2814 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2816 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2817 advertise SMTP AUTH mechanism to us, instead of a generic
2818 protocol violation error. Also, make Exim more robust to bad
2819 data from the Dovecot auth socket.
2821 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2823 When a queue runner is handling a message, Exim first routes the
2824 recipient addresses, during which it prunes them based on the retry
2825 hints database. After that it attempts to deliver the message to
2826 any remaining recipients. It then updates the hints database using
2829 So if a recipient address works intermittently, it can get repeatedly
2830 deferred at routing time. The retry hints record remains fresh so the
2831 address never reaches the final cutoff time.
2833 This is a fairly common occurrence when a user is bumping up against
2834 their storage quota. Exim had some logic in its local delivery code
2835 to deal with this. However it did not apply to per-recipient defers
2836 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2838 This change adds a proper retry rule check during routing so that the
2839 final cutoff time is checked against the message's age. We only do
2840 this check if there is an address retry record and there is not a
2841 domain retry record; this implies that previous attempts to handle
2842 the address had the retry_use_local_parts option turned on. We use
2843 this as an approximation for the destination being like a local
2844 delivery, as in LMTP.
2846 I suspect this new check makes the old local delivery cutoff check
2847 redundant, but I have not verified this so I left the code in place.
2849 TF/02 Correct gecos expansion when From: is a prefix of the username.
2851 Test 0254 submits a message to Exim with the header
2855 When I ran the test suite under the user fanf2, Exim expanded
2856 the header to contain my full name, whereas it should have added
2857 a Resent-Sender: header. It erroneously treats any prefix of the
2858 username as equal to the username.
2860 This change corrects that bug.
2862 GF/01 DCC debug and logging tidyup
2863 Error conditions log to paniclog rather than rejectlog.
2864 Debug lines prefixed by "DCC: " to remove any ambiguity.
2866 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2868 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2869 Bug spotted by Jeremy Harris; was flawed since initial commit.
2870 Would have resulted in OCSP responses post-SNI triggering an Exim
2871 NULL dereference and crash.
2873 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2875 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2876 Bug detection, analysis and fix by Samuel Thibault.
2877 Bugzilla 1331, Debian bug #698092.
2879 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2881 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2882 Server implementation by Todd Lyons, client by JH.
2883 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2884 config variable "prdr_enable" controls whether the server
2885 advertises the facility. If the client requests PRDR a new
2886 acl_data_smtp_prdr ACL is called once for each recipient, after
2887 the body content is received and before the acl_smtp_data ACL.
2888 The client is controlled by both of: a hosts_try_prdr option
2889 on the smtp transport, and the server advertisement.
2890 Default client logging of deliveries and rejections involving
2891 PRDR are flagged with the string "PRDR".
2893 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2894 fclose(). Diagnosis by Todd Lyons.
2896 PP/17 Update configure.default to handle IPv6 localhost better.
2897 Patch by Alain Williams (plus minor tweaks).
2900 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2901 This is now consistent with GnuTLS, and is now documented: the
2902 previous undocumented portable approach to treating the option as
2903 unset was to force an expansion failure. That still works, and
2904 an empty string is now equivalent.
2906 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2907 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2908 not performing validation itself.
2910 PP/20 Added force_command boolean option to pipe transport.
2911 Patch from Nick Koston, of cPanel Inc.
2913 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2916 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2918 PP/21 Fix eximon continuous updating with timestamped log-files.
2919 Broken in a format-string cleanup in 4.80, missed when I repaired the
2920 other false fix of the same issue.
2921 Report and fix from Heiko Schlichting.
2924 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2925 Report from Prashanth Katuri.
2927 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2928 It's SecureTransport, so affects any MacOS clients which use the
2929 system-integrated TLS libraries, including email clients.
2931 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2932 using a MIME ACL for non-SMTP local injection.
2933 Report and assistance in diagnosis by Warren Baker.
2935 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2937 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2939 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2940 last id that failed so it may be referenced in subsequent ACL's.
2942 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2945 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2946 ldap library initialization, allowing self-signed CA's to be
2947 used. Also properly sets require_cert option later in code by
2948 using NULL (global ldap config) instead of ldap handle (per
2949 session). Bug diagnosis and testing by alxgomz.
2951 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2952 the src/util/ subdirectory.
2954 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2955 renamed to Transport Post Delivery Action by Jeremy Harris, as
2958 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2959 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2960 redis_servers = needs to be configured which will be used by the redis
2961 lookup. Patch from Warren Baker, of The Packet Hub.
2963 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2965 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2966 hostname or reverse DNS when processing a host list. Used suggestions
2967 from multiple comments on this bug.
2969 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2971 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2972 Missed a few lines, added it to make the runtest require no keyboard
2975 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2976 contains upper case chars. Make router use caseful_local_part.
2978 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2979 support when GnuTLS has been built with p11-kit.
2985 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2987 This, or similar/improved, will also be change PP/11 of 4.82.
2993 PP/01 Handle short writes when writing local log-files.
2994 In practice, only affects FreeBSD (8 onwards).
2995 Bugzilla 1053, with thanks to Dmitry Isaikin.
2997 NM/01 Bugzilla 949 - Documentation tweak
2999 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
3002 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
3004 PP/02 Implemented gsasl authenticator.
3006 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
3008 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
3009 `pkg-config foo` for cflags/libs.
3011 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
3012 with rest of GSASL and with heimdal_gssapi.
3014 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
3015 `pkg-config foo` for cflags/libs for the TLS implementation.
3017 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
3018 properties get this fed in as external SSF. A number of robustness
3019 and debugging improvements to the cyrus_sasl authenticator.
3021 PP/08 cyrus_sasl server now expands the server_realm option.
3023 PP/09 Bugzilla 1214 - Log authentication information in reject log.
3024 Patch by Jeremy Harris.
3026 PP/10 Added dbmjz lookup type.
3028 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
3030 PP/12 MAIL args handles TAB as well as SP, for better interop with
3031 non-compliant senders.
3032 Analysis and variant patch by Todd Lyons.
3034 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
3035 Bug report from Lars Müller <lars@samba.org> (via SUSE),
3036 Patch from Dirk Mueller <dmueller@suse.com>
3038 PP/13 tls_peerdn now print-escaped for spool files.
3039 Observed some $tls_peerdn in wild which contained \n, which resulted
3040 in spool file corruption.
3042 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
3043 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
3044 or write after TLS renegotiation, which otherwise led to messages
3047 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3048 as a tracking header (ie: a signed header comes before the signature).
3049 Patch from Wolfgang Breyha.
3051 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3052 comma-sep list; embedded commas doubled.
3054 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3056 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3058 Report and patch from Dmitry Banschikov.
3060 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3061 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3062 failures. We appear to now support TLS1.1+ with Exim.
3064 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3065 lets Exim select keys and certificates based upon TLS SNI from client.
3066 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3067 before an outbound SMTP session. New log_selector, +tls_sni.
3069 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3070 NULL dereference. Report and patch from Alun Jones.
3072 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3073 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3074 Not seeing resolver debug output on NetBSD, but suspect this is a
3075 resolver implementation change.
3077 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3078 Left warnings. Added "eximon gdb" invocation mode.
3080 PP/21 Defaulting "accept_8bitmime" to true, not false.
3082 PP/22 Added -bw for inetd wait mode support.
3084 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3085 locate the relevant includes and libraries. Made this the default.
3087 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3088 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3090 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3091 This may cause build issues on older platforms.
3093 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3094 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3095 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3096 Added SNI support via GnuTLS too.
3097 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3099 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3101 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3102 Applied second patch from Janne, implementing suggestion to default
3103 multiple-strings-in-record handling to match SPF spec.
3105 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3107 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3108 read-only, out of scope).
3109 Patch from Wolfgang Breyha, report from Stuart Northfield.
3111 PP/29 Fix three issues highlighted by clang analyser static analysis.
3112 Only crash-plausible issue would require the Cambridge-specific
3113 iplookup router and a misconfiguration.
3114 Report from Marcin Mirosław.
3116 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3118 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3119 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3120 As part of this, removing so much warning spew let me fix some minor
3121 real issues in debug logging.
3123 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3124 assignment on my part. Fixed.
3126 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3127 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3128 Janne Snabb (who went above and beyond: thank you).
3130 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3131 string otherwise requires a connection and a bunch more work and it's
3132 relatively easy to get wrong. Should also expose TLS library linkage
3135 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3136 64-bit ${eval} (JH/03).
3138 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3139 GNU libc to support some of the 64-bit stuff, should not lead to
3140 conflicts. Defined before os.h is pulled in, so if a given platform
3141 needs to override this, it can.
3143 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3144 protection layer was required, which is not implemented.
3145 Bugzilla 1254, patch from Wolfgang Breyha.
3147 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3148 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3149 tls_dhparam take prime identifiers. Also unbreak combination of
3150 OpenSSL+DH_params+TLSSNI.
3152 PP/39 Disable SSLv2 by default in OpenSSL support.
3158 PP/01 Solaris build fix for Oracle's LDAP libraries.
3159 Bugzilla 1109, patch from Stephen Usher.
3161 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3163 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3166 TF/02 Fix a couple more cases where we did not log the error message
3167 when unlink() failed. See also change 4.74-TF/03.
3169 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3170 lock up or crash if it happened to be inside a call to libc when it
3171 got a SIGUSR1 from exiwhat.
3173 The SIGUSR1 handler appends the current process status to the process
3174 log which is later printed by exiwhat. It used to use the general
3175 purpose logging code to do this, but several functions it calls are
3176 not safe for signals.
3178 The new output code in the SIGUSR1 handler is specific to the process
3179 log, and simple enough that it's easy to inspect for signal safety.
3180 Removing some special cases also simplifies the general logging code.
3181 Removing the spurious timestamps from the process log simplifies
3184 TF/04 Improved ratelimit ACL condition.
3186 The /noupdate option has been deprecated in favour of /readonly which
3187 has clearer semantics. The /leaky, /strict, and /readonly update modes
3188 are mutually exclusive. The update mode is no longer included in the
3189 database key; it just determines when the database is updated. (This
3190 means that when you upgrade Exim will forget old rate measurements.)
3192 Exim now checks that the per_* options are used with an update mode that
3193 makes sense for the current ACL. For example, when Exim is processing a
3194 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3195 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3196 must specify per_mail/readonly. If you omit the update mode it defaults to
3197 /leaky where that makes sense (as before) or /readonly where required.
3199 The /noupdate option is now undocumented but still supported for
3200 backwards compatibility. It is equivalent to /readonly except that in
3201 ACLs where /readonly is required you may specify /leaky/noupdate or
3202 /strict/noupdate which are treated the same as /readonly.
3204 A useful new feature is the /count= option. This is a generalization
3205 of the per_byte option, so that you can measure the throughput of other
3206 aggregate values. For example, the per_byte option is now equivalent
3207 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3209 The per_rcpt option has been generalized using the /count= mechanism
3210 (though it's more complicated than the per_byte equivalence). When it is
3211 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3212 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3213 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3214 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3215 Note that using per_rcpt with a non-readonly update mode in more than
3216 one ACL will cause the recipients to be double-counted. (The per_mail
3217 and per_byte options don't have this problem.)
3219 The handling of very low rates has changed slightly. If the computed rate
3220 is less than the event's count (usually one) then this event is the first
3221 after a long gap. In this case the rate is set to the same as this event's
3222 count, so that the first message of a spam run is counted properly.
3224 The major new feature is a mechanism for counting the rate of unique
3225 events. The new per_addr option counts the number of different
3226 recipients that someone has sent messages to in the last time period. It
3227 behaves like per_rcpt if all the recipient addresses are different, but
3228 duplicate recipient addresses do not increase the measured rate. Like
3229 the /count= option this is a general mechanism, so the per_addr option
3230 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3231 example, measure the rate that a client uses different sender addresses
3232 with the options per_mail/unique=$sender_address. There are further
3233 details in the main documentation.
3235 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3237 TF/06 Removed a few PCRE remnants.
3239 TF/07 Automatically extract Exim's version number from tags in the git
3240 repository when doing development or release builds.
3242 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3243 Bugzilla 879. Patch from Paul Fisher.
3245 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3246 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3249 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3251 PP/05 Variable $av_failed, true if the AV scanner deferred.
3252 Bugzilla 1078. Patch from John Horne.
3254 PP/06 Stop make process more reliably on build failure.
3255 Bugzilla 1087. Patch from Heiko Schlittermann.
3257 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3258 Bugzilla 1089. Patch from Heiko Schlittermann.
3260 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3261 Bugzilla 1131. Patch from Holger Weiß.
3263 PP/09 Handle IPv6 addresses with SPF.
3264 Bugzilla 860. Patch from Wolfgang Breyha.
3266 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3268 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3271 PP/11 match_* no longer expand right-hand-side by default.
3272 New compile-time build option, EXPAND_LISTMATCH_RHS.
3273 New expansion conditions, "inlist", "inlisti".
3275 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3277 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3279 PP/14 fix log_write() format string regression from TF/03.
3280 Bugzilla 1152. Patch from Dmitry Isaikin.
3286 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3288 PP/02 Harmonised TLS library version reporting; only show if debugging.
3289 Layout now matches that introduced for other libraries in 4.74 PP/03.
3291 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3293 PP/04 New "dns_use_edns0" global option.
3295 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3298 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3299 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3301 TK/01 Updated PolarSSL code to 0.14.2.
3302 Bugzilla 1097. Patch from Andreas Metzler.
3304 PP/07 Catch divide-by-zero in ${eval:...}.
3305 Fixes bugzilla 1102.
3307 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3310 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3311 format-string attack -- SECURITY: remote arbitrary code execution.
3313 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3314 time unintentionally subject to list matching rules, letting the header
3315 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3316 arbitrary string expansion). This allowed for information disclosure.
3318 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3319 INT_MIN/-1 -- value coerced to INT_MAX.
3325 NM/01 Workaround for PCRE version dependency in version reporting
3328 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3329 This fixes portability to compilers other than gcc, notably
3330 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3332 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3333 makefiles for portability to HP-UX and POSIX correctness.
3335 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3336 Also via indented variable definition in the Makefile.
3337 (Debugging by Oliver Heesakkers).
3339 PP/02 Restore caching of spamd results with expanded spamd_address.
3340 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3342 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3343 Improves build reliability. Fix from: Frank Elsner
3345 NM/02 Fix wide character breakage in the rfc2047 coding
3346 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3348 NM/03 Allow underscore in dnslist lookups
3349 Fixes bug 1026. Patch from Graeme Fowler
3351 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3352 Code patches from Adam Ciarcinski of NetBSD.
3354 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3357 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3358 is logged, to avoid truncation. Patch from John Horne.
3360 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3361 Patch from Jakob Hirsch.
3363 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3364 SQL string expansion failure details.
3365 Patch from Andrey Oktyabrski.
3367 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3368 Patch from Simon Arlott.
3370 PP/09 New lookups functionality failed to compile on old gcc which rejects
3371 extern declarations in function scope.
3372 Patch from Oliver Fleischmann
3374 PP/10 Use sig_atomic_t for flags set from signal handlers.
3375 Check getgroups() return and improve debugging.
3376 Fixed developed for diagnosis in bug 927 (which turned out to be
3379 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3380 Patch from Mark Zealey.
3382 PP/12 Bugzilla 1056: Improved spamd server selection.
3383 Patch from Mark Zealey.
3385 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3386 Based on patch from Heiko Schlittermann.
3388 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3389 Patch from Uwe Doering, sign-off by Michael Haardt.
3391 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3392 variable declaration deep within a block. Bug and patch from
3395 PP/15 lookups-Makefile IRIX compatibility coercion.
3397 PP/16 Make DISABLE_DKIM build knob functional.
3399 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3400 Patch by Simon Arlott
3402 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3403 variable argument macros. Our copy now differs from upstream.
3409 TF/01 Failure to get a lock on a hints database can have serious
3410 consequences so log it to the panic log.
3412 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3413 controlled using the smtp_confirmation log selector.
3415 TF/03 Include the error message when we fail to unlink a spool file.
3417 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3418 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3419 for maintaining out-of-tree patches for some time.
3421 PP/01 Bugzilla 139: Documentation and portability issues.
3422 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3423 Handle per-OS dynamic-module compilation flags.
3425 PP/02 Let /dev/null have normal permissions.
3426 The 4.73 fixes were a little too stringent and complained about the
3427 permissions on /dev/null. Exempt it from some checks.
3428 Reported by Andreas M. Kirchwitz.
3430 PP/03 Report version information for many libraries, including
3431 Exim version information for dynamically loaded libraries. Created
3432 version.h, now support a version extension string for distributors
3433 who patch heavily. Dynamic module ABI change.
3435 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3436 privilege escalation vulnerability whereby the Exim run-time user
3437 can cause root to append content of the attacker's choosing to
3440 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3443 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3444 If dropping privileges for untrusted macros, we disabled normal logging
3445 on the basis that it would fail; for the Exim run-time user, this is not
3446 the case, and it resulted in successful deliveries going unlogged.
3447 Fixed. Reported by Andreas Metzler.
3453 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3454 only prepend for the Resent-* case. Fixes regression introduced in
3455 Exim 4.70 by NM/22 for Bugzilla 607.
3457 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3458 increasing numbers of administrators be bitten by this.
3460 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3462 PP/03 Bugzilla 994: added openssl_options main configuration option.
3464 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3466 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3468 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3470 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3471 without a peer certificate, leading to a segfault because of an
3472 assumption that peers always have certificates. Be a little more
3473 paranoid. Problem reported by Martin Tscholak.
3475 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3476 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3477 NB: ClamAV planning to remove STREAM in "middle of 2010".
3478 CL also introduces -bmalware, various -d+acl logging additions and
3479 more caution in buffer sizes.
3481 PP/09 Implemented reverse_ip expansion operator.
3483 PP/10 Bugzilla 937: provide a "debug" ACL control.
3485 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3487 PP/12 Bugzilla 973: Implement --version.
3489 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3491 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3493 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3495 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3496 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3497 ignore trailing whitespace.
3499 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3501 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3504 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3505 Notification from Dr Andrew Aitchison.
3507 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3508 ExtendedDetectionInfo response format.
3509 Notification from John Horne.
3511 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3514 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3515 XSL and documented dependency on system catalogs, with examples of how
3518 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3521 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3522 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3523 configuration file which is writeable by the Exim user or group.
3525 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3526 of configuration files to cover files specified with the -C option if
3527 they are going to be used with root privileges, not just the default
3530 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3531 option (effectively making it always true).
3533 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3534 files to be used while preserving root privileges.
3536 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3537 that rogue child processes cannot use them.
3539 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3540 run-time user, instead of root.
3542 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3543 Exim run-time user without dropping privileges.
3545 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3546 result string, instead of calling string_vformat() twice with the same
3549 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3550 for other users. Others should always drop root privileges if they use
3551 -C on the command line, even for a whitelisted configure file.
3553 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3555 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3561 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3562 $data_path, and $header_path variables; fixed documentation bugs and
3565 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3566 exipick to access non-standard spools, including the "frozen" queue
3569 NM/01 Bugzilla 965: Support mysql stored procedures.
3570 Patch from Alain Williams
3572 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3574 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3575 Patch from Andreas Metzler
3577 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3578 Patch from Kirill Miazine
3580 NM/05 Bugzilla 671: Added umask to procmail example.
3582 JJ/03 installed exipick 20100323.0, fixing doc bug
3584 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3585 directory. Notification and patch from Dan Rosenberg.
3587 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3589 TK/02 Improve log output when DKIM signing operation fails.
3591 MH/01 Treat the transport option dkim_domain as a colon separated
3592 list, not as a single string, and sign the message with each element,
3593 omitting multiple occurences of the same signer.
3595 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3596 Bugzilla 985, 986. Patch by Simon Arlott
3598 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3599 Patch by Simon Arlott
3601 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3602 MBX locking. Notification from Dan Rosenberg.
3608 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3610 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3612 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3614 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3616 NM/04 Bugzilla 915: Flag broken perl installation during build.
3622 TK/01 Added patch by Johannes Berg that expands the main option
3623 "spamd_address" if it starts with a dollar sign.
3625 TK/02 Write list of recipients to X-Envelope-Sender header when building
3626 the mbox-format spool file for content scanning (suggested by Jakob
3629 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3630 (http://www.dcc-servers.net/) support via dccifd. Activated by
3631 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3633 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3634 by Mark Daniel Reidel <mr@df.eu>.
3636 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3637 When building exim an external PCRE library is now needed -
3638 PCRE is a system library on the majority of modern systems.
3639 See entry on PCRE_LIBS in EDITME file.
3641 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3642 conversation. Added nologin parameter to request.
3643 Patch contributed by Kirill Miazine.
3645 TF/01 Do not log submission mode rewrites if they do not change the address.
3647 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3649 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3650 log files in place. Contributed by Roberto Lima.
3652 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3654 TF/03 Bugzilla 615: When checking the local_parts router precondition
3655 after a local_part_suffix or local_part_prefix option, Exim now
3656 does not use the address's named list lookup cache, since this
3657 contains cached lookups for the whole local part.
3659 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3660 Robert Millan. Documentation is in experimental-spec.txt.
3662 TF/04 Bugzilla 668: Fix parallel build (make -j).
3664 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3666 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3667 Patch provided by Jan Srzednicki.
3669 TF/05 Leading white space used to be stripped from $spam_report which
3670 wrecked the formatting. Now it is preserved.
3672 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3673 that they are available at delivery time.
3675 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3677 TF/08 TLS error reporting now respects the incoming_interface and
3678 incoming_port log selectors.
3680 TF/09 Produce a more useful error message if an SMTP transport's hosts
3681 setting expands to an empty string.
3683 NM/06 Bugzilla 744: EXPN did not work under TLS.
3684 Patch provided by Phil Pennock.
3686 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3687 Patch provided by Richard Godbee.
3689 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3690 acl_smtp_notquit, added index entry.
3692 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3693 Patch provided by Eugene Bujak.
3695 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3696 accept(). Patch provided by Maxim Dounin.
3698 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3699 Patch provided by Phil Pennock.
3701 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3703 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3704 Patch provided by Brad "anomie" Jorsch.
3706 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3707 Patch provided by Dean Brooks.
3709 TK/05 Add native DKIM support (does not depend on external libraries).
3711 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3712 Patch provided by Graeme Fowler.
3714 NM/16 Bugzilla 851: Documentation example syntax fix.
3716 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3718 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3721 NM/19 Bugzilla 745: TLS version reporting.
3722 Patch provided by Phil Pennock.
3724 NM/20 Bugzilla 167: bool: condition support.
3725 Patch provided by Phil Pennock.
3727 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3728 clients. Patch provided by Phil Pennock.
3730 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3731 Patch provided by Brad "anomie" Jorsch.
3733 NM/23 Bugzilla 687: Fix misparses in eximstats.
3734 Patch provided by Heiko Schlittermann.
3736 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3737 Patch provided by Heiko Schlittermann.
3739 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3740 plus update to original patch.
3742 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3744 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3745 Patch provided by David Brownlee.
3747 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3749 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3751 NM/30 Bugzilla 888: TLS documentation bugfixes.
3753 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3755 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3756 Unlike the original bugzilla I have changed all shell scripts in src tree.
3758 NM/33 Bugzilla 898: Transport filter timeout fix.
3759 Patch by Todd Rinaldo.
3761 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3762 Patch by Serge Demonchaux.
3764 NM/35 Bugzilla 39: Base64 decode bug fixes.
3765 Patch by Jakob Hirsch.
3767 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3769 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3771 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3773 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3779 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3780 ALT-N's libdkim that I have put here:
3781 http://duncanthrax.net/exim-experimental/
3783 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3784 were called 'true' and it seems that C99 defines that as a reserved
3785 keyword to be used with 'bool' variable types. That means you could
3786 not include C99-style headers which use bools without triggering
3787 build errors in sieve.c.
3789 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3790 as mailq or other aliases. Changed the --help handling significantly
3791 to do whats expected. exim_usage() emits usage/help information.
3793 SC/01 Added the -bylocaldomain option to eximstats.
3795 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3797 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3799 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3805 PH/01 Another patch from the Sieve maintainer.
3807 PH/02 When an IPv6 address is converted to a string for single-key lookup
3808 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3809 dots are used instead of colons so that keys in lsearch files need not
3810 contain colons. This was done some time before quoting was made available
3811 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3812 (notated using the quote facility) so as to distinguish them from IPv4
3813 keys. This meant that lookups for IP addresses in host lists did not work
3814 for iplsearch lookups.
3816 This has been fixed by arranging for IPv6 addresses to be expressed with
3817 colons if the lookup type is iplsearch. This is not incompatible, because
3818 previously such lookups could never work.
3820 The situation is now rather anomalous, since one *can* have colons in
3821 ordinary lsearch keys. However, making the change in all cases is
3822 incompatible and would probably break a number of configurations.
3824 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3827 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3828 conversion specification without a maximum field width, thereby enabling
3829 a rogue spamd server to cause a buffer overflow. While nobody in their
3830 right mind would setup Exim to query an untrusted spamd server, an
3831 attacker that gains access to a server running spamd could potentially
3832 exploit this vulnerability to run arbitrary code as the Exim user.
3834 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3835 $primary_hostname instead of what libspf2 thinks the hosts name is.
3837 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3838 a directory entry by the name of the lookup key. Previously, if a
3839 symlink pointed to a non-existing file or a file in a directory that
3840 Exim lacked permissions to read, a lookup for a key matching that
3841 symlink would fail. Now it is enough that a matching directory entry
3842 exists, symlink or not. (Bugzilla 503.)
3844 PH/03 The body_linecount and body_zerocount variables are now exported in the
3847 PH/04 Added the $dnslist_matched variable.
3849 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3850 This means they are set thereafter only if the connection becomes
3853 PH/06 Added the client_condition to authenticators so that some can be skipped
3854 by clients under certain conditions.
3856 PH/07 The error message for a badly-placed control=no_multiline_responses left
3857 "_responses" off the end of the name.
3859 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3861 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3862 (without spaces) instead of just copying the configuration text.
3864 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3866 PH/11 Added $max_received_linelength.
3868 PH/12 Added +ignore_defer and +include_defer to host lists.
3870 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3871 way in which PCRE > 7.0 is built.
3873 PH/14 Implemented queue_only_load_latch.
3875 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3876 MAIL command. The effect was to mangle the value on 64-bit systems.
3878 PH/16 Another patch from the Sieve maintainer.
3880 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3882 PH/18 If a system quota error occurred while trying to create the file for
3883 a maildir delivery, the message "Mailbox is full" was not appended to the
3884 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3885 only to a quota excession during the actual writing of the file.
3887 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3888 characters?) which causes problems in log lines. The DN values are now
3889 passed through string_printing() before being added to log lines.
3891 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3892 and InterBase are left for another time.)
3894 PH/21 Added message_body_newlines option.
3896 PH/22 Guard against possible overflow in moan_check_errorcopy().
3898 PH/23 POSIX allows open() to be a macro; guard against that.
3900 PH/24 If the recipient of an error message contained an @ in the local part
3901 (suitably quoted, of course), incorrect values were put in $domain and
3902 $local_part during the evaluation of errors_copy.
3908 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3909 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3912 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3913 issue a MAIL command.
3915 PH/02 In an ACL statement such as
3917 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3919 if a client was not listed at all, or was listed with a value other than
3920 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3921 the condition was not true (as it should be), so access was not denied.
3922 The bug was that the ! inversion was incorrectly passed on to the second
3923 item. This has been fixed.
3925 PH/03 Added additional dnslists conditions == and =& which are different from
3926 = and & when the dns lookup returns more than one IP address.
3928 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3929 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3931 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3932 FSYNC, which compiles an option called disable_fsync that allows for
3933 bypassing fsync(). The documentation is heavily laced with warnings.
3935 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3937 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3938 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3939 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3940 including adding "make clean"; (3) Added -fPIC when compiling the test
3941 dynamically loaded module, to get rid of a warning.
3943 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3944 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3945 = 0s. The bug is otherwise harmless.
3947 PH/07 There was a bug in the dovecot authenticator such that the value of
3948 $auth1 could be overwritten, and so not correctly preserved, after a
3949 successful authentication. This usually meant that the value preserved by
3950 the server_setid option was incorrect.
3952 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3954 PH/09 Installed PCRE release 7.0.
3956 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3957 run for batched SMTP input. It is now run at the start of every message
3958 in the batch. While fixing this I discovered that the process information
3959 (output by running exiwhat) was not always getting set for -bs and -bS
3960 input. This is fixed, and it now also says "batched" for BSMTP.
3962 PH/11 Added control=no_pipelining.
3964 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3965 patch, slightly modified), and move the expansion of helo_data till after
3966 the connection is made in the smtp transport (so it can use these
3969 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3971 PH/14 Added log_selector = +pid.
3973 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3975 PH/16 Add ${if forany and ${if forall.
3977 PH/17 Added dsn_from option to vary the From: line in DSNs.
3979 PH/18 Flush SMTP output before performing a callout, unless control =
3980 no_callout_flush is set.
3982 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3983 was true (the default) a successful delivery failed to delete the retry
3984 item, thus causing premature timeout of the address. The bug is now
3987 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3989 PH/21 Long custom messages for fakedefer and fakereject are now split up
3990 into multiline responses in the same way that messages for "deny" and
3991 other ACL rejections are.
3993 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3994 with slight modification.
3996 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3997 draft, changing the syntax and factoring some duplicate code".
3999 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
4000 for deliveries of the second and subsequent messages over the same SMTP
4003 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
4004 ${reduce, with only minor "tidies".
4006 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
4008 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
4009 expansion side effects.
4011 PH/27 When a message times out after an over-quota error from an Exim-imposed
4012 quota, the bounce message says "mailbox is full". This message was not
4013 being given when it was a system quota that was exceeded. It now should
4016 MH/03 Made $recipients available in local_scan(). local_scan() already has
4017 better access to the recipient list through recipients_list[], but
4018 $recipients can be useful in postmaster-provided expansion strings.
4020 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
4021 in the case of a MAIL command with additional options following the
4022 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
4023 were accidentally chopped off.
4025 PH/29 SMTP synchronization checks are implemented when a command is read -
4026 there is a check that no more input is waiting when there shouldn't be
4027 any. However, for some commands, a delay in an ACL can mean that it is
4028 some time before the response is written. In this time, more input might
4029 arrive, invalidly. So now there are extra checks after an ACL has run for
4030 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
4031 pipelining has not been advertised.
4033 PH/30 MH's patch to allow iscntrl() characters to be list separators.
4035 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
4036 returned in the SMTP response when smtp_return_error_details was false.
4037 This has been fixed.
4039 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
4040 instead of the C I/O that was originally supplied, because problems were
4041 reported on Solaris.
4043 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
4044 Exim which did not show up earlier: it was assuming that a call to
4045 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4046 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4047 was a macro that became an assignment, so it seemed to work. This has
4048 changed to a proper function call with a void return, hence the compile
4049 error. Exim's code has been fixed.
4051 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4054 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4056 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4058 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4059 because it turns out to be too big or there is a timeout) there is no
4060 "Completed" line in the log. When some messages of this type were
4061 selected by exigrep, they were listed as "not completed". Others were
4062 picked up by some special patterns. I have improved the selection
4063 criteria to be more general.
4065 PH/38 The host_find_failed option in the manualroute router can now be set
4066 to "ignore", to completely ignore a host whose IP address cannot be
4067 found. If all hosts are ignored, the behaviour is controlled by the new
4068 host_all_ignored option.
4070 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4071 homing or because of multiple MX records with /mx) generated more than
4072 one IP address, and the following item turned out to be the local host,
4073 all the secondary addresses of the first item were incorrectly removed
4074 from the list, along with the local host and any following hosts (which
4075 is what is supposed to happen).
4077 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4078 whoever called Exim into the -H file. In the case of the daemon it was
4079 behaving confusingly. When first started, it used values for whoever
4080 started the daemon, but after a SIGHUP it used the Exim user (because it
4081 calls itself on a restart). I have changed the code so that it now always
4084 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4085 message are rejected with the same error (e.g. no authentication or bad
4086 sender address), and a DATA command is nevertheless sent (as can happen
4087 with PIPELINING or a stupid MUA), the error message that was given to the
4088 RCPT commands is included in the rejection of the DATA command. This is
4089 intended to be helpful for MUAs that show only the final error to their
4092 PH/42 Another patch from the Sieve maintainer.
4094 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4095 Eximstats - Fixed some broken HTML links and added missing column headers
4097 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4098 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4100 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4102 PH/43 Yet another patch from the Sieve maintainer.
4104 PH/44 I found a way to check for a TCP/IP connection going away before sending
4105 the response to the final '.' that terminates a message, but only in the
4106 case where the client has not sent further data following the '.'
4107 (unfortunately, this is allowed). However, in many cases there won't be
4108 any further data because there won't be any more messages to send. A call
4109 to select() can be used: if it shows that the input is "ready", there is
4110 either input waiting, or the socket has been closed. An attempt to read
4111 the next input character can distinguish the two cases. Previously, Exim
4112 would have sent an OK response which the client would never have see.
4113 This could lead to message repetition. This fix should cure that, at
4114 least in a lot of common cases.
4116 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4117 advertised in response to EHLO.
4123 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4124 fixed by 4.65/MH/01 (is this a record?) are fixed:
4126 (i) An empty string was always treated as zero by the numeric comparison
4127 operators. This behaviour has been restored.
4129 (ii) It is documented that the numeric comparison operators always treat
4130 their arguments as decimal numbers. This was broken in that numbers
4131 starting with 0 were being interpreted as octal.
4133 While fixing these problems I realized that there was another issue that
4134 hadn't been noticed. Values of message_size_limit (both the global option
4135 and the transport option) were treated as octal if they started with 0.
4136 The documentation was vague. These values are now always treated as
4137 decimal, and I will make that clear in the documentation.
4143 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4144 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4147 MH/01 Don't check that the operands of numeric comparison operators are
4148 integers when their expansion is in "skipping" mode (fixes bug
4149 introduced by 4.64-PH/07).
4151 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4152 child addresses, Exim now panics and dies. Previously, because the count
4153 is held in a short int, deliveries were likely to be lost. As such a
4154 large number of recipients for a single message is ridiculous
4155 (performance will be very, very poor), I have chosen to impose a limit
4156 rather than extend the field.
4162 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4163 leftover -K file (the existence of which was triggered by #402).
4164 While we were at it, introduced process PID as part of the -K
4165 filename. This should rule out race conditions when creating
4168 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4169 processing considerably. Previous code took too long for large mails,
4170 triggering a timeout which in turn triggers #401.
4172 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4173 in the DK code in transports.c. sendfile() is not really portable,
4174 hence the _LINUX specificness.
4176 TF/01 In the add_headers option to the mail command in an Exim filter,
4177 there was a bug that Exim would claim a syntax error in any
4178 header after the first one which had an odd number of characters
4181 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4182 callout verification, Exim cached a "reject" for the entire domain. This
4183 is correct for most verifications, but it is not correct for a recipient
4184 verification with use_sender or use_postmaster set, because in that case
4185 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4186 case of MAIL FROM:<> rejection from other early rejections (e.g.
4187 rejection of HELO). When verifying a recipient using a non-null MAIL
4188 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4189 Whatever the result of the callout, the value of the domain cache is
4190 left unchanged (for any other kind of callout, getting as far as trying
4191 RCPT means that the domain itself is ok).
4193 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4196 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4197 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4198 session, the daemon ignores ECONNECT errors and logs others; it now
4199 ignores EPIPE as well.
4201 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4202 (quoted-printable decoding).
4204 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4205 later the small subsequent patch to fix an introduced bug.
4207 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4209 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4211 PH/08 An error is now given if message_size_limit is specified negative.
4213 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4214 to be given (somewhat) arbitrary names.
4216 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4219 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4220 miscellaneous code fixes
4222 PH/10 Added the log_reject_target ACL modifier to specify where to log
4225 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4226 hostname. This is wrong, because it relates to the incoming message (and
4227 probably the interface on which it is arriving) and not to the outgoing
4228 callout (which could be using a different interface). This has been
4229 changed to use the value of the helo_data option from the smtp transport
4230 instead - this is what is used when a message is actually being sent. If
4231 there is no remote transport (possible with a router that sets up host
4232 addresses), $smtp_active_hostname is used.
4234 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4235 tweaks were necessary in order to get it to work (see also 21 below):
4236 (a) The code assumed that strncpy() returns a negative number on buffer
4237 overflow, which isn't the case. Replaced with Exim's string_format()
4239 (b) There were several signed/unsigned issues. I just did the minimum
4240 hacking in of casts. There is scope for a larger refactoring.
4241 (c) The code used strcasecmp() which is not a standard C function.
4242 Replaced with Exim's strcmpic() function.
4243 (d) The code set only $1; it now sets $auth1 as well.
4244 (e) A simple test gave the error "authentication client didn't specify
4245 service in request". It would seem that Dovecot has changed its
4246 interface. Fortunately there's a specification; I followed it and
4247 changed what the client sends and it appears to be working now.
4249 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4252 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4253 address A is aliased to B and C, where B exists and C does not. Without
4254 -v the output is "A verified" because verification stops after a
4255 successful redirection if more than one address is generated. However,
4256 with -v the child addresses are also verified. Exim was outputting "A
4257 failed to verify" and then showing the successful verification for C,
4258 with its parentage. It now outputs "B failed to verify", showing B's
4259 parentage before showing the successful verification of C.
4261 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4262 look up a TXT record in a specific list after matching in a combined
4265 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4266 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4267 they consult the DNS. I had assumed they would set it the way they
4268 wanted; and indeed my experiments on Linux seem to show that in some
4269 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4270 To be on the safe side, however, I have now made the interface to
4271 host_find_byname() similar to host_find_bydns(), with an argument
4272 containing the DNS resolver options. The host_find_byname() function now
4273 sets these options at its start, just as host_find_bydns() does. The smtp
4274 transport options dns_qualify_single and dns_search_parents are passed to
4275 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4276 of host_find_byname() use the default settings of RES_DEFNAMES
4277 (qualify_single) but not RES_DNSRCH (search_parents).
4279 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4280 spool_read_header() do less string testing, by means of a preliminary
4281 switch on the second character of optional "-foo" lines. (This is
4282 overdue, caused by the large number of possibilities that now exist.
4283 Originally there were few.) While I was there, I also converted the
4284 str(n)cmp tests so they don't re-test the leading "-" and the first
4285 character, in the hope this might squeeze out yet more improvement.
4287 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4288 flag allowing group syntax was set by the header_syntax check but not
4289 turned off, possible causing trouble later; (2) The flag was not being
4290 set at all for the header_verify test, causing "group"-style headers to
4291 be rejected. I have now set it in this case, and also caused header_
4292 verify to ignore an empty address taken from a group. While doing this, I
4293 came across some other cases where the code for allowing group syntax
4294 while scanning a header line wasn't quite right (mostly, not resetting
4295 the flag correctly in the right place). These bugs could have caused
4296 trouble for malformed header lines. I hope it is now all correct.
4298 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4299 with the "reply" argument non-NULL. The code, however (which originally
4300 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4301 but it didn't always do it. This confused somebody who was copying the
4302 code for some other use. I have removed all the tests.
4304 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4305 feature that was used to support insecure browsers during the U.S. crypto
4306 embargo. It requires special client support, and Exim is probably the
4307 only MTA that supported it -- and would never use it because real RSA is
4308 always available. This code has been removed, because it had the bad
4309 effect of slowing Exim down by computing (never used) parameters for the
4310 RSA_EXPORT functionality.
4312 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4313 authenticator to fail if there's a tab character in the incoming data
4314 (there should never be unless someone is messing about, as it's supposed
4315 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4316 if the connection is using TLS or if the remote IP is the same as the
4317 local IP, and the "valid-client-cert option" if a client certificate has
4320 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4321 authenticators. This can be used for authorization after authentication
4322 succeeds. (In the case of plaintext, it servers for both authentication
4325 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4326 if any retry times were supplied.
4328 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4329 connection, where there is no client IP address to check. In this
4330 situation, the verify now always succeeds.
4332 PH/25 Applied John Jetmore's -Mset patch.
4334 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4336 PH/27 In a string expansion for a processed (not raw) header when multiple
4337 headers of the same name were present, leading whitespace was being
4338 removed from all of them, but trailing whitespace was being removed only
4339 from the last one. Now trailing whitespace is removed from each header
4340 before concatenation. Completely empty headers in a concatenation (as
4341 before) are ignored.
4343 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4344 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4346 PH/29 [Removed. This was a change that I later backed out, and forgot to
4347 correct the ChangeLog entry (that I had efficiently created) before
4348 committing the later change.]
4350 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4351 address errors (4xx response to RCPT) over the same connection as other
4352 messages routed to the same hosts. Such deliveries are always "forced",
4353 so retry times are not inspected. This resulted in far too many retries
4354 for the affected addresses. The effect occurred only when there were more
4355 hosts than the hosts_max_try setting in the smtp transport when it had
4356 the 4xx errors. Those hosts that it had tried were not added to the list
4357 of hosts for which the message was waiting, so if all were tried, there
4358 was no problem. Two fixes have been applied:
4360 (i) If there are any address or message errors in an SMTP delivery, none
4361 of the hosts (tried or untried) are now added to the list of hosts
4362 for which the message is waiting, so the message should not be a
4363 candidate for sending over the same connection that was used for a
4364 successful delivery of some other message. This seems entirely
4365 reasonable: after all the message is NOT "waiting for some host".
4366 This is so "obvious" that I'm not sure why it wasn't done
4367 previously. Hope I haven't missed anything, but it can't do any
4368 harm, as the worst effect is to miss an optimization.
4370 (ii) If, despite (i), such a delivery is accidentally attempted, the
4371 routing retry time is respected, so at least it doesn't keep
4372 hammering the server.
4374 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4375 in ${readsocket because some servers need this prod.
4377 PH/32 Added some extra debug output when updating a wait-xxx database.
4379 PH/33 The hint "could be header name not terminated by colon", which has been
4380 given for certain expansion errors for a long time, was not being given
4381 for the ${if def:h_colon_omitted{... case.
4383 PH/34 The spec says: "With one important exception, whenever a domain list is
4384 being scanned, $domain contains the subject domain." There was at least
4385 one case where this was not true.
4387 PH/35 The error "getsockname() failed: connection reset by peer" was being
4388 written to the panic log as well as the main log, but it isn't really
4389 panic-worthy as it just means the connection died rather early on. I have
4390 removed the panic log writing for the ECONNRESET error when getsockname()
4393 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4394 runs only) independently of the message's sender address. This meant
4395 that, if the 4xx error was in fact related to the sender, a different
4396 message to the same recipient with a different sender could confuse
4397 things. In particular, this can happen when sending to a greylisting
4398 server, but other circumstances could also provoke similar problems.
4399 I have changed the default so that the retry time for these errors is now
4400 based a combination of the sender and recipient addresses. This change
4401 can be overridden by setting address_retry_include_sender=false in the
4404 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4405 remote server are returned as part of bounce messages. This was not
4406 happening for LMTP over a pipe (the lmtp transport), but now it is the
4407 same for both kinds of LMTP.
4409 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4410 in header lines that were in fact CNAMEs. This is no longer the case.
4412 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4413 and queue runs started by the daemon processed all messages. This has
4414 been fixed so that -R and -S can now usefully be given with -q<time>.
4416 PH/40 Import PCRE release 6.7 (fixes some bugs).
4418 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4420 PH/42 Give an error if -q is specified more than once.
4422 PH/43 Renamed the variables $interface_address and $interface_port as
4423 $received_ip_address and $received_port, to make it clear that these
4424 values apply to message reception, and not to the outgoing interface when
4425 a message is delivered. (The old names remain recognized, of course.)
4427 PH/44 There was no timeout on the connect() call when using a Unix domain
4428 socket in the ${readsocket expansion. There now is.
4430 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4431 be meaningful with "accept".
4433 SC/01 Eximstats V1.43
4434 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4436 SC/02 Eximstats V1.44
4437 Use a glob alias rather than an array ref in the generated
4438 parser. This improves both readability and performance.
4440 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4441 Collect SpamAssassin and rejection statistics.
4442 Don't display local sender or destination tables unless
4443 there is data to show.
4444 Added average volumes into the top table text output.
4446 SC/04 Eximstats V1.46
4447 Collect data on the number of addresses (recipients)
4448 as well as the number of messages.
4450 SC/05 Eximstats V1.47
4451 Added 'Message too big' to the list of mail rejection
4452 reasons (thanks to Marco Gaiarin).
4454 SC/06 Eximstats V1.48
4455 Mainlog lines which have GMT offsets and are too short to
4456 have a flag are now skipped.
4458 SC/07 Eximstats V1.49 (Alain Williams)
4459 Added the -emptyok flag.
4461 SC/08 Eximstats V1.50
4462 Fixes for obtaining the IP address from reject messages.
4464 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4465 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4466 whitespace changes from 4.64-PH/27
4468 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4471 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4472 are found, allow negative numbers in numeric criteria)
4474 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4476 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4477 to match changes made in 4.64-PH/43
4479 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4481 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4482 "domain" is actually an IP address, and force a failure. This locks out
4483 those revolvers/nameservers that support "A-for-A" lookups, in
4484 contravention of the specifications.
4486 PH/48 When a host name was looked up from an IP address, and the subsequent
4487 forward lookup of the name timed out, the host name was left in
4488 $sender_host_name, contrary to the specification.
4490 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4491 restricted to single-key lookups, Exim was not diagnosing an error if
4492 * or *@ was used with a query-style lookup.
4494 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4496 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4497 long ago, but noone interested enough thought of it. Let's just say that
4498 the "1.1" means that there are some new functions that weren't there at
4499 some point in the past.
4501 PH/51 Error processing for expansion failure of helo_data from an smtp
4502 transport during callout processing was broken.
4504 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4505 tested/used via the -bh/-bhc/-bs options.
4507 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4508 bug, fixed in subsequent PCRE releases).
4510 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4511 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4513 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4519 SC/01 Use a glob alias rather than an array ref in eximstats generated
4520 parser. This improves both readability and performance.
4522 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4523 Don't display local sender or destination tables in eximstats unless
4524 there is data to show.
4525 Added average volumes into the eximstats top table text output.
4527 SC/03 Collect data on the number of addresses (recipients) as well
4528 as the number of messages in eximstats.
4530 TF/01 Correct an error in the documentation for the redirect router. Exim
4531 does (usually) call initgroups() when daemonizing.
4533 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4534 with consistent privilege compared to when running as a daemon.
4536 TF/03 Note in the spec that $authenticated_id is not set for local
4537 submissions from trusted users.
4539 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4540 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4542 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4543 by adding some example configuration directives to the default
4544 configuration file. A little bit of work is required to uncomment the
4545 directives and define how usernames and passwords are checked, but
4546 there is now a framework to start from.
4548 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4549 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4550 without this. I don't know how relevant this is to other LDAP libraries.
4552 PH/02 Add the verb name to the "unknown ACL verb" error.
4554 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4556 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4558 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4559 directory not expanded when it should be if an expanded home directory
4560 was set for the address (which is overridden by the transport).
4562 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4565 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4566 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4567 because it is too late at that time, and has no effect.
4569 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4570 security issue with \' (bugzilla #107). I could not use the
4571 PQescapeStringConn() function, because it needs a PGconn value as one of
4574 PH/08 When testing addresses using -bt, indicate those final addresses that
4575 are duplicates that would not cause an additional delivery. At least one
4576 person was confused, thinking that -bt output corresponded to deliveries.
4577 (Suppressing duplicates isn't a good idea as you lose the information
4578 about possibly different redirections that led to the duplicates.)
4580 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4581 systems where poll() doesn't work, in particular OS X.
4583 PH/10 Added more information to debugging output for retry time not reached.
4585 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4586 operations in malware.c.
4588 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4591 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4592 syslog_duplication set false, log lines that would normally be written
4593 both the the main log and to the reject log were not written to syslog at
4596 PH/14 In the default configuration, change the use of "message" in ACL warn
4597 statements to "add_header".
4599 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4600 not followed by a command (e.g. "seen endif").
4602 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4603 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4606 PH/17 Added extra conditions to the default value of delay_warning_condition
4610 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4611 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4612 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4615 The Auto-Submitted: and various List- headers are standardised, whereas I
4616 don't think Precedence: ever was.
4618 PH/18 Refactored debugging code in route_finduser() to show more information,
4619 in particular, the error code if getpwnam() issues one.
4621 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4622 This is apparently needed in addition to the PH/07 change above to avoid
4623 any possible encoding problems.
4625 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4626 but not after initializing Perl.
4628 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4629 output them only if debugging. By default they are written stderr,
4630 apparently, which is not desirable.
4632 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4635 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4638 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4640 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4641 authenticated or an ident call has been made. Suppress the default
4642 values for $authenticated_id and $authenticated_sender (but permit -oMai
4643 and -oMas) when testing with -bh.
4645 PH/24 Re-jigged the order of the tests in the default configuration so that the
4646 tests for valid domains and recipients precede the DNS black list and CSA
4647 tests, on the grounds that those ones are more expensive.
4649 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4650 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4651 This bug exists in every version of Exim that I still have, right back to
4654 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4655 However, an attempt to turn on case-sensitivity in a regex key by
4656 including (?-i) didn't work because the subject string was already
4657 lowercased, and the effects were non-intuitive. It turns out that a
4658 one-line patch can be used to allow (?-i) to work as expected.
4664 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4665 other effects) broke the use of negated acl sub-conditions.
4667 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4670 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4671 "Deny" causes Exim to reject the incoming connection with a 554 error.
4672 Unfortunately, if there is a major crisis, such as a disk failure,
4673 tcp-wrappers gives "deny", whereas what one would like would be some
4674 kind of temporary error. A kludge has been added to help with this.
4675 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4676 554 error is used if errno is still zero or contains ENOENT (which occurs
4677 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4680 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4682 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4683 errors. Otherwise a message that provokes a temporary error (when other
4684 messages do not) can cause a whole host to time out.
4686 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4687 addresses were routed directly to files or pipes from a redirect router.
4688 File deliveries just didn't batch; pipe deliveries might have suffered
4691 PH/06 A failure to get a lock for a hints database would erroneously always say
4692 "Failed to get write lock", even when it was really a read lock.
4694 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4695 of 0600. This has been changed to use the value of the lockfile_mode
4696 option (which defaults to 0600).
4698 PH/08 Applied small patch from the Sieve maintainer.
4700 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4701 folder from quota calculations, a direct delivery into this folder messed
4702 up the contents of the maildirsize file. This was because the regex was
4703 used only to exclude .Trash (or whatever) when the size of the mailbox
4704 was calculated. There was no check that a delivery was happening into an
4705 excluded directory. This bug has been fixed by ignoring all quota
4706 processing for deliveries into excluded directories.
4708 PH/10 Added the maildirfolder_create_regex option to appendfile.
4714 PH/01 The code for finding all the local interface addresses on a FreeBSD
4715 system running IPv6 was broken. This may well have applied to all BSD
4716 systems, as well as to others that have similar system calls. The broken
4717 code found IPv4 interfaces correctly, but gave incorrect values for the
4718 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4719 that it would not match correctly against @[] and not recognize the IPv6
4722 PH/02 The ipliteral router was not recognizing addresses of the form user@
4723 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4725 PH/03 Added disable_ipv6.
4727 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4728 decoded form, because it is most often used to construct To: headers
4729 lines in autoreplies, and the decoded form may well be syntactically
4730 invalid. However, $reply_address has leading white space removed, and all
4731 newlines turned into spaces so that the autoreply transport does not
4734 PH/05 If group was specified without a user on a router, and no group or user
4735 was specified on a transport, the group from the router was ignored.
4737 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4738 for visible compile-time settings that can be used to change these
4739 numbers, for those that want even more. Backwards compatibility with old
4740 spool files has been maintained. However, going back to a previous Exim
4741 release will lost any variables that are in spool files.
4743 PH/07 Two small changes when running in the test harness: increase delay when
4744 passing a TCP/IP connection to a new process, in case the original
4745 process has to generate a bounce, and remove special handling of
4746 127.0.0.2 (sic), which is no longer necessary.
4748 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4749 be the same on different OS.
4751 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4754 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4755 whether --show-vars was specified or not
4757 JJ/02 exipick: Added support for new ACL variable spool format introduced
4760 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4761 syntactically invalid From: or Reply-to: line, and a filter used this to
4762 generate an autoreply, and therefore failed to obtain an address for the
4763 autoreply, Exim could try to deliver to a non-existent relative file
4764 name, causing unrelated and misleading errors. What now happens is that
4765 it logs this as a hard delivery error, but does not attempt to create a
4768 PH/11 The exinext utility has a -C option for testing purposes, but although
4769 the given file was scanned by exinext itself; it wasn't being passed on
4770 when Exim was called.
4772 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4773 an end-of-file indication when reading a command response.
4775 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4776 compiled. In many other places in Exim, IPv6 addresses are always
4777 recognized, so I have changed this. It also means that IPv4 domain
4778 literals of the form [IPV4:n.n.n.n] are now always recognized.
4780 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4781 used if the router is not running as root, for example, when verifying at
4782 ACL time, or when using -bh. The debugging output from this situation was
4783 non-existent - all you got was a failure to exec. I have made two
4786 (a) Failures to set uid/gid, the current directory, or a process leader
4787 in a subprocess such as that created by queryprogram now generate
4788 suitable debugging output when -d is set.
4790 (b) The queryprogram router detects when it is not running as root,
4791 outputs suitable debugging information if -d is set, and then runs
4792 the subprocess without attempting to change uid/gid.
4794 PH/15 Minor change to Makefile for building test_host (undocumented testing
4797 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4798 additional section of a DNS packet that returns MX or SRV records.
4799 Instead, it always explicitly searches for A/AAAA records. This avoids
4800 major problems that occur when a DNS server includes only records of one
4801 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4802 fixed another bug: if SRV records were looked up and the corresponding
4803 address records were *not* found in the additional section, the port
4804 values from the SRV records were lost.
4806 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4807 using the correct key (the original address) when searching the retry
4808 rules in order to find which one to use for generating the retry hint.
4810 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4811 adding the default one. Similarly, if it contains a Reply-To: header, the
4812 errors_reply_to option, if set, is not used.
4814 PH/19 When calculating a retry time, Exim used to measure the "time since
4815 failure" by looking at the "first failed" field in the retry record. Now
4816 it does not use this if it is later than than the arrival time of the
4817 message. Instead it uses the arrival time. This makes for better
4818 behaviour in cases where some deliveries succeed, thus re-setting the
4819 "first failed" field. An example is a quota failure for a huge message
4820 when small messages continue to be delivered. Without this change, the
4821 "time since failure" will always be short, possible causing more frequent
4822 delivery attempts for the huge message than are intended.
4823 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4825 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4826 $1, $2, $3) because the numerical variables can be reset during some
4827 expansion items (e.g. "match"), thereby losing the authentication data.
4829 PH/21 Make -bV show the size of off_t variables so that the test suite can
4830 decide whether to run tests for quotas > 2G.
4832 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4833 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4834 filecount value is greater than 2G or if a quota value is greater than 2G
4835 on a system where the size of off_t is not greater than 4, a panic error
4838 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4839 never match. The debug and -bh output now contains an explicit error
4840 message indicating a malformed IPv4 address or mask.
4842 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4843 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4844 PH/23 above applies.
4846 PH/25 Do not write to syslog when running in the test harness. The only
4847 occasion when this arises is a failure to open the main or panic logs
4848 (for which there is an explicit test).
4850 PH/26 Added the /no_tell option to "control=freeze".
4852 PH/27 If a host name lookup failed very early in a connection, for example, if
4853 the IP address matched host_lookup and the reverse lookup yielded a name
4854 that did not have a forward lookup, an error message of the form "no IP
4855 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4856 could be logged. Now it outputs the IP address instead of "NULL".
4858 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4859 allows the sender and the authenticated sender to be set when
4860 submitting a message from within Exim. Since child_open_exim() is
4861 documented for local_scan(), the new function should be too.
4863 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4864 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4865 results in an empty string is now treated as unset.
4867 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4869 PH/31 Added sender_verify_fail logging option.
4871 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4872 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4873 tidied the source and removed it altogether.
4875 PH/33 When a queue run was abandoned because the load average was too high, a
4876 log line was always written; now it is written only if the queue_run log
4877 selector is set. In addition, the log line for abandonment now contains
4878 information about the queue run such as the pid. This is always present
4879 in "start" and "stop" lines but was omitted from the "abandon" line.
4881 PH/34 Omit spaces between a header name and the colon in the error message that
4882 is given when verify = headers_syntax fails (if there are lots of them,
4883 the message gets confusing).
4885 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4886 names, as there are now some PTR records that contain slashes. This check
4887 is only to protect against broken name servers that fall over on strange
4888 characters, so the fact that it applies to all lookups doesn't matter.
4890 PH/36 Now that the new test suite is complete, we can remove some of the
4891 special code in Exim that was needed for the old test suite. For example,
4892 sorting DNS records because real resolvers return them in an arbitrary
4893 order. The new test suite's fake resolver always returns records in the
4896 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4897 bounces) except when queue_only is set, to avoid logging races between
4898 the different processes.
4900 PH/38 Panic-die if .include specifies a non-absolute path.
4902 PH/39 A tweak to the "H" retry rule from its user.
4904 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4905 a label. They prevented compilation on older perls.
4907 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4908 a warning to be raised on newish perls.
4910 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4911 on queue. Changes to match documented behaviour of showing count of
4912 messages matching specified criteria.
4914 PH/40 Changed the default ident timeout from 30s to 5s.
4916 PH/41 Added support for the use of login_cap features, on those BSD systems
4917 that have them, for controlling the resources used by pipe deliveries.
4919 PH/42 The content-scanning code uses fopen() to create files in which to put
4920 message data. Previously it was not paying any attention to the mode of
4921 the files. Exim runs with umask(0) because the rest of the code creates
4922 files with open(), and sets the required mode explicitly. Thus, these
4923 files were ending up world-writeable. This was not a big issue, because,
4924 being within the spool directory, they were not world-accessible. I have
4925 created a function called modefopen, which takes an additional mode
4926 argument. It sets umask(777), creates the file, chmods it to the required
4927 mode, then resets the umask. All the relevant calls to fopen() in the
4928 content scanning code have been changed to use this function.
4930 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4931 to 24 hours. This avoids potential overflow problems when processing G
4932 and H retry rules. I suspect nobody ever tinkers with this value.
4934 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4936 PH/45 When the plaintext authenticator is running as a client, the server's
4937 challenges are checked to ensure they are valid base64 strings. By
4938 default, the authentication attempt is cancelled if an invalid string is
4939 received. Setting client_ignore_invalid_base64 true ignores these errors.
4940 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4941 they are received. Thus, the responses can be made to depend on the
4942 challenges. If an invalid string is ignored, an empty string is placed in
4945 PH/46 Messages that are created by the autoreply transport now contains a
4946 References: header, in accordance with RFCs 2822 and 3834.
4948 PH/47 Added authenticated_sender_force to the smtp transport.
4950 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4952 PH/49 Installed latest patch from the Sieve maintainer.
4954 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4955 was also set, the appendfile transport was unnecessarily scanning a
4956 directory of message files (e.g. for maildir delivery) to find the count
4957 of files (along with the size), even though it did not need this
4958 information. It now does the scan only if it needs to find either the
4959 size of the count of files.
4961 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4963 PH/52 Two bugs concerned with error handling when the smtp transport is
4966 (i) Exim was not creating retry information for temporary errors given
4967 for individual recipients after the DATA command when the smtp transport
4968 was used in LMTP mode. This meant that they could be retried too
4969 frequently, and not timed out correctly.
4971 (ii) Exim was setting the flag that allows error details to be returned
4972 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4973 recipients that were returned after the DATA command.
4975 PH/53 This is related to PH/52, but is more general: for any failing address,
4976 when detailed error information was permitted to be returned to the
4977 sender, but the error was temporary, then after the final timeout, only
4978 "retry timeout exceeded" was returned. Now it returns the full error as
4979 well as "retry timeout exceeded".
4981 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4982 do this, and (what is worse) MTAs that accept it.
4984 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4985 will now be deprecated.
4987 PH/56 New os.c-cygwin from the Cygwin maintainer.
4989 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4990 formats (previously only available in exim formats via -bpr, -bpru,
4991 and -bpra. Now also available in native and exiqgrep formats)
4993 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4994 with very large, slow to parse queues
4996 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4998 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
5000 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
5001 responses to authentication challenges, though it was showing the
5002 challenges; (ii) I've removed the CR characters from the debug output for
5005 PH/58 Allow for the insertion of a newline as well as a space when a string
5006 is turned into more than one encoded-word during RFC 2047 encoding. The
5007 Sieve code now uses this.
5009 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
5010 data_4xx, lost_connection, tls_required.
5012 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
5013 message was being sent as an SMTP response.
5015 PH/61 Add -l and -k options to exicyclog.
5017 PH/62 When verifying, if an address was redirected to one new address, so that
5018 verification continued, and the new address failed or deferred after
5019 having set something in $address_data, the value of $address_data was not
5020 passed back to the ACL. This was different to the case when no
5021 redirection occurred. The value is now passed back in both cases.
5023 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
5024 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
5025 use login_cap.h, so on its own it isn't the distinguishing feature. The
5026 new name refers directly to the setclassresources() function.
5028 PH/65 Added configuration files for NetBSD3.
5030 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
5032 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
5033 is preferred over IPv4.
5035 PH/68 The bounce_return_message and bounce_return_body options were not being
5036 honoured for bounces generated during the reception of non-SMTP messages.
5037 In particular, this applied to messages rejected by the ACL. This bug has
5038 been fixed. However, if bounce_return_message is true and bounce_return_
5039 body is false, the headers that are returned for a non-SMTP message
5040 include only those that have been read before the error was detected.
5041 (In the case of an ACL rejection, they have all been read.)
5043 PH/69 The HTML version of the specification is now built in a directory called
5044 spec_html instead of spec.html, because the latter looks like a path with
5045 a MIME-type, and this confuses some software.
5047 PH/70 Catch two compiler warnings in sieve.c.
5049 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5050 function verify_get_ident() calls ip_connect() to connect a socket, but
5051 if the "connect()" function timed out, ip_connect() used to close the
5052 socket. However, verify_get_ident() also closes the socket later, and in
5053 between Exim writes to the log, which may get opened at this point. When
5054 the socket was closed in ip_connect(), the log could get the same file
5055 descriptor number as the socket. This naturally causes chaos. The fix is
5056 not to close the socket in ip_connect(); the socket should be closed by
5057 the function that creates it. There was only one place in the code where
5058 this was missing, in the iplookup router, which I don't think anybody now
5059 uses, but I've fixed it anyway.
5061 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5062 well as to direct DNS lookups. Otherwise the handling of names in host
5063 lists is inconsistent and therefore confusing.
5069 PH/01 Two changes to the default runtime configuration:
5071 (1) Move the checks for relay_from_hosts and authenticated clients from
5072 after to before the (commented out) DNS black list checks.
5074 (2) Add control=submission to the relay_from_hosts and authenticated
5075 clients checks, on the grounds that messages accepted by these
5076 statements are most likely to be submissions.
5078 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5080 (1) Generate an error if the third argument for the ${prvs expansion is
5083 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5086 (3) Reset the variables that are obtained from the first argument of
5087 ${prvscheck and used in the second argument before leaving the code,
5088 because their memory is reclaimed, so using them afterwards may do
5091 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5092 one (it's much easier than Tom thought :-).
5094 (5) Because of (4), we can now allow for the use of $prvscheck_result
5095 inside the third argument.
5097 PH/03 For some reason, the default setting of PATH when running a command from
5098 a pipe transport was just "/usr/bin". I have changed it to
5101 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5102 anything to be listed in the output from -bV.
5104 PH/05 When a filter generated an autoreply, the entire To: header line was
5105 quoted in the delivery log line, like this:
5107 => >A.N.Other <ano@some.domain> <original@ddress> ...
5109 This has been changed so that it extracts the operative address. There
5110 may be more than one such address. If so, they are comma-separated, like
5113 => >ano@some.domain,ona@other.domain <original@ddress> ...
5115 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5116 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5117 not being looked up in the rDNS to get a host name, Exim was showing the
5118 IP address twice in Received: lines, even though the IP addresses were
5119 identical. For example:
5121 Received: from [1.2.3.4] (helo=[1.2.3.4])
5123 However, if the real host name was known, it was omitting the HELO data
5124 if it matched the actual IP address. This has been tidied up so that it
5125 doesn't show the same IP address twice.
5127 PH/07 When both +timestamp and +memory debugging was on, the value given by
5128 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5129 called by the debug printing, thereby overwriting the timestamp buffer.
5130 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5132 PH/08 When the original message was included in an autoreply transport, it
5133 always said "this is a copy of the message, including all the headers",
5134 even if body_only or headers_only was set. It now gives an appropriate
5137 PH/09 Applied a patch from the Sieve maintainer which:
5139 o fixes some comments
5140 o adds the (disabled) notify extension core
5141 o adds some debug output for the result of if/elsif tests
5142 o points to the current vacation draft in the documentation
5143 and documents the missing references header update
5147 o fixes a bug in processing the envelope test (when testing
5148 multiple envelope elements, the last element determined the
5151 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5152 Electronic Mail") by including:
5154 Auto-submitted: auto-generated
5156 in the messages that it generates (bounce messages and others, such as
5157 warnings). In the case of bounce messages for non-SMTP messages, there was
5158 also a typo: it was using "Auto_submitted" (underscore instead of
5159 hyphen). Since every message generated by Exim is necessarily in response
5160 to another message, thes have all been changed to:
5162 Auto-Submitted: auto-replied
5164 in accordance with these statements in the RFC:
5166 The auto-replied keyword:
5168 - SHOULD be used on messages sent in direct response to another
5169 message by an automatic process,
5171 - MUST NOT be used on manually-generated messages,
5173 - MAY be used on Delivery Status Notifications (DSNs) and Message
5174 Disposition Notifications (MDNs),
5176 - MUST NOT be used on messages generated by automatic or periodic
5177 processes, except for messages which are automatic responses to
5180 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5181 to the default Received: header definition.
5183 PH/12 Added log selector acl_warn_skipped (default on).
5185 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5186 variables because (a) they are in the wrong storage pool and (b) even if
5187 they were copied, it wouldn't work properly because of the caching.
5189 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5190 checking when decoding. Apparently there are clients that generate
5191 overlong encoded strings. Why am I not surprised?
5193 PH/15 If the first argument of "${if match_address" was not empty, but did not
5194 contain an "@" character, Exim crashed. Now it writes a panic log message
5195 and treats the condition as false.
5197 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5199 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5200 extension "envelope-auth". The code is finished and in agreement with
5201 other implementations, but there is no documentation so far and in fact,
5202 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5203 not changing the active code.
5205 Print executed "if" and "elsif" statements when debugging is used. This
5206 helps a great deal to understand what a filter does.
5208 Document more things not specified clearly in RFC3028. I had all this
5209 sorted out, when out of a sudden new issues came to my mind. Oops."
5211 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5214 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5215 canonical form (as documented). However, after a host name lookup from
5216 the IP address, check_host() was doing a simple string comparison with
5217 addresses acquired from the DNS when checking that the found name did
5218 have the original IP as one of its addresses. Since any found IPv6
5219 addresses are likely to be in abbreviated form, the comparison could
5220 fail. Luckily, there already exists a function for doing the comparison
5221 by converting both addresses to binary, so now that is used instead of
5222 the text comparison.
5224 PH/20 There was another similar case to PH/19, when a complete host name was
5225 given in a host list; looking up its IP address could give an abbreviated
5226 form, whereas the current host's name might or might not be abbreviated.
5227 The same fix has been applied.
5233 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5234 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5237 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5238 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5240 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5242 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5243 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5244 Also, for "Auto-Submitted", treat anything other than "no" as
5245 non-personal, in accordance with RFC 3834. (Previously it treated
5246 anything starting "auto-" as non-personal.)
5248 TF/01 The control=submission/name=... option had a problem with syntax
5249 errors if the name included a slash character. The /name= option
5250 now slurps the rest of the string, so it can include any characters
5251 but it must come last in the list of options (after /sender_retain
5254 PH/05 Some modifications to the interface to the fake nameserver for the new
5262 TK/01 Added the "success_on_redirect" address verification option. See
5263 NewStuff for rationale and an example.
5265 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5267 PH/02 Patch to exigrep to allow it to work on syslog lines.
5269 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5270 fread() to skip over the body file's header line, because in Cygwin the
5271 header line is locked and is inaccessible.
5273 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5274 co-exist for some time) to make it clear that it is the Exim ID that is
5275 referenced, not the Message-ID: header line.
5277 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5278 string_format() function, because snprintf() does not exist on all
5281 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5282 ${stat: expansion item.
5284 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5285 protocol synchronization error", to keep the pedants happy.
5287 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5288 well as for IRIX systems, when gcc is being used. See the host.c source
5291 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5293 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5296 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5297 transport and to the smtp transport in LMTP mode.
5299 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5301 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5302 run a filter in a subprocess. This could lead to confusion in subsequent
5303 lookups in the parent process. There should also be a search_tidyup() at
5304 the end of the subprocess.
5306 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5307 only if the host matched helo_try_verify_hosts, which caused the
5308 verification to occur when the EHLO/HELO command was issued. The ACL just
5309 tested the remembered result. Now, if a previous verification attempt has
5310 not happened, "verify = helo" does it there and then.
5312 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5314 TK/03 Fix log output including CR from clamd.
5316 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5317 exist provoked a memory error which could cause a segfault.
5319 PH/15 Installed PCRE 6.2
5321 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5323 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5324 of the problem. Specifically, suggested +O2 rather than +O1 for the
5327 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5329 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5330 record, the port was not correctly logged when the outgoing_port log
5331 selector was set (it logged the transort's default port).
5333 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5334 fallback_hosts, and "hosts" in the smtp transport.
5336 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5337 host errors such as "Connection refused".
5339 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5340 authentication with radiusclient 0.4.9:
5342 - Error returned from rc_read_config was caught wrongly
5343 - Username/password not passed on to radius server due to wrong length.
5345 The presumption is that some radiusclient API changes for 4.51/PH/17
5346 were not taken care of correctly. The code is still untested by me (my
5347 Linux distribution still has 0.3.2 of radiusclient), but it was
5348 contributed by a Radius user.
5350 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5351 expanding the "port" option of the smtp transport.
5353 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5354 while reading a MIME header. Thanks to Tom Hughes for a patch.
5356 PH/24 Include config.h inside local_scan.h so that configuration settings are
5359 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5360 that in an ACL for RCPT (for example), you can examine exactly what was
5363 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5364 commands, but it was not correctly comparing the address with the actual
5365 client host address. Thus, it would show the EHLO address in Received:
5366 header lines when this was not necessary.
5368 PH/27 Added the % operator to ${eval:}.
5370 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5371 it should be ignoring failures (because with -C, for example, it has lost
5372 privilege). It wasn't ignoring creation failures other than "already
5375 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5378 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5379 because an input line was too long, either on its own, or by virtue of
5380 too many continuations, the temporary file was not being removed, and the
5381 return code was incorrect.
5383 PH/31 Missing "BOOL" in function definition in filtertest.c.
5385 PH/32 Applied Sieve patches from the maintainer.
5387 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5389 PH/33 Added "verify = not_blind".
5391 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5392 Local/Makefile (with some defaults set). These are used in built scripts
5393 such as exicyclog, but they have never been used in the exim_install
5394 script (though there are many overriding facilities there). I have
5395 arranged that the exim_install script now takes note of these two
5398 PH/35 Installed configuration files for Dragonfly.
5400 PH/36 When a locally submitted message by a trusted user did not contain a
5401 From: header, and the sender address was obtained from -f or from an SMTP
5402 MAIL command, and the trusted user did not use -F to supply a sender
5403 name, $originator_name was incorrectly used when constructing a From:
5404 header. Furthermore, $originator_name was used for submission mode
5405 messages from external hosts without From: headers in a similar way,
5406 which is clearly wrong.
5408 PH/37 Added control=suppress_local_fixups.
5410 PH/38 When log_selector = +received_sender was set, and the addition of the
5411 sender made the log line's construction buffer exactly full, or one byte
5412 less than full, an overflow happened when the terminating "\n" was
5415 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5416 when the result of a list match is failure because a DNS lookup failed.
5418 PH/40 RM_COMMAND is now used in the building process.
5420 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5421 the "build-* directories that it finds.
5423 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5424 address in a domain literal was a prefix of an interface address.
5426 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5427 when verifying a sender address, unless rewrite_headers is false.
5429 PH/44 Wrote a long comment about why errors_to addresses are verified as
5430 recipients, not senders.
5432 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5433 the ratelimit ACL was added.
5435 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5437 PH/46 Added extra information about PostgreSQL errors to the error string.
5439 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5440 avoiding the need to install special zones in a real server. This is
5441 backwards compatible; if it can't find the fake resolver, it drops back.
5442 Thus, both old and new test suites can be run.
5444 TF/02 Added util/ratelimit.pl
5446 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5447 clock is set back in time.
5449 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5450 Candler <B.Candler@pobox.com>.
5452 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5453 OK for addresses that are the result of redirections.
5455 PH/48 A number of further additions for the benefit of the new test suite,
5456 including a fake gethostbyname() that interfaces to the fake DNS resolver
5459 TF/06 The fix for widen_domains has also been applied to qualify_single and
5460 search_parents which are the other dnslookup options that can cause
5463 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5466 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5468 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5469 service" if it encountered an MX record with an empty target hostname.
5470 The message is now "an MX or SRV record indicated no SMTP service".
5472 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5473 if the DNS of the sending site is misconfigured. This is quite a
5474 common situation. This change restores the behaviour of treating a
5475 helo verification defer as a failure.
5477 PH/51 If self=fail was set on a router, the bounce message did not include the
5478 actual error message.
5484 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5486 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5487 command itself ended in error, the underlying message about the transport
5488 filter timeout was being overwritten with the pipe command error. Now the
5489 underlying error message should be appended to the second error message.
5491 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5493 PH/02 Reduce the amount of output that "make" produces by default. Full output
5494 can still be requested.
5496 PH/03 The warning log line about a condition test deferring for a "warn" verb
5497 was being output only once per connection, rather than after each
5498 occurrence (because it was using the same function as for successful
5499 "warn" verbs). This seems wrong, so I have changed it.
5501 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5502 it should not have, which might have caused a crash in the right
5503 circumstances, but probably never did.
5505 PH/04 Installed a modified version of Tony Finch's patch to make submission
5506 mode fix the return path as well as the Sender: header line, and to
5507 add a /name= option so that you can make the user's friendly name appear
5510 TF/03 Added the control = fakedefer ACL modifier.
5512 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5513 Mark Lowes for thorough testing.
5515 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5517 TK/03 Merged latest SRS patch from Miles Wilton.
5519 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5520 with the definition in sysexits.h (which is #included earlier).
5521 Fortunately, Exim does not actually use EX_OK. The code used to try to
5522 preserve the sysexits.h value, by assuming that macro definitions were
5523 scanned for macro replacements. I have been disabused of this notion,
5524 so now the code just undefines EX_OK before #including unistd.h.
5526 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5527 in the smtp transport. When a block could not be written in a single
5528 write() function, the timeout was being re-applied to each part-write.
5529 This seems wrong - if the receiver was accepting one byte at a time it
5530 would take for ever. The timeout is now adjusted when this happens. It
5531 doesn't have to be particularly precise.
5533 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5534 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5536 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5537 without a domain if the check to <postmaster@domain> fails.
5539 SC/01 Eximstats: added -xls and the ability to specify output files
5540 (patch written by Frank Heydlauf).
5542 SC/02 Eximstats: use FileHandles for outputting results.
5544 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5546 SC/04 Eximstats: fixed display of large numbers with -nvr option
5548 SC/05 Eximstats: fixed merging of reports with empty tables.
5550 SC/06 Eximstats: added the -include_original_destination flag
5552 SC/07 Eximstats: removed tabs and trailing whitespace.
5554 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5556 TK/06 MBOX spool code: Add real "From " MBOX separator line
5557 so the .eml file is really in mbox format (even though
5558 most programs do not really care). Patch from Alex Miller.
5560 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5561 The latter is generated from $received_to and is only set if the
5562 message has one envelope recipient. SA can use these headers,
5563 obviously out-of-the-box. Patch from Alex Miller.
5565 PH/08 The ${def test on a variable was returning false if the variable's
5566 value was "0", contrary to what the specification has always said!
5567 The result should be true unless the variable is empty.
5569 PH/09 The syntax error of a character other than { following "${if
5570 def:variable_name" (after optional whitespace) was not being diagnosed.
5571 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5572 accidental colon was present, for example, could give incorrect results.
5574 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5575 result is used (not including appendfile, where other changes are about
5578 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5579 This involved changing a lot of size variables from int to off_t. It
5580 should work with maildirs and everything.
5582 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5583 spamd dying while we are connected to it.
5585 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5588 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5589 function for BDB 4.3.
5591 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5593 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5594 the sources for PCRE have been re-arranged and more files are now
5597 PH/15 The code I had for printing potentially long long variables in PH/11
5598 above was not the best (it lost precision). The length of off_t variables
5599 is now inspected at build time, and an appropriate printing format (%ld
5600 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5601 to be "long long int" or "long int". This is needed for the internal
5602 formatting function string_vformat().
5604 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5605 the configuration file to be ":syslog", then the script "guesses" where
5606 the logs files are, rather than using the compiled in default. In our
5607 case the guess is not the same as the compiled default, so the script
5608 suddenly stopped working when I started to use syslog. The patch checks
5609 to see if log_file_path is "". If so, it attempts to read it from exim
5610 with no configuration file to get the compiled in version, before it
5611 falls back to the previous guessing code."
5613 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5614 implementing BATV in an Exim configuration. See NewStuff for the gory
5617 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5618 Makefile that are specific to HP-UX.
5620 PH/18 If the "use_postmaster" option was set for a recipient callout together
5621 with the "random" option, the postmaster address was used as the MAIL
5622 FROM address for the random test, but not for the subsequent recipient
5623 test. It is now used for both.
5625 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5626 patch removes a few documentation additions to RFC 3028, because the
5627 latest draft now contains them. It adds the new en;ascii-case comparator
5628 and a new error check for 8bit text in MIME parts. Comparator and
5629 require names are now matched exactly. I enabled the subaddress
5630 extension, but it is not well tested yet (read: it works for me)."
5632 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5633 rework some of the code of TK/09 above to avoid the hardwired use of
5634 "%lld" and "long long". Replaced the call to snprintf() with a call to
5637 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5638 records point to non-existent hosts", "retry timeout exceeded", and
5639 "retry time not reached for any host after a long failure period".
5641 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5642 experimental DomainKeys support:
5644 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5645 (2) On an error such as an illegally used "control", the wrong name for
5646 the control was given.
5648 These problems did NOT occur unless DomainKeys support was compiled.
5650 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5652 PH/24 Added ${if match_ip condition.
5654 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5655 databases so that it will be absolutely obvious if a crash occurs in the
5656 DB library. This is a regular occurrence (often caused by mis-matched
5659 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5660 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5661 picked up on a user's system that detects such things. There doesn't seem
5662 to be a gcc warning option for this - only an attribute that has to be
5663 put on the function's prototype. It seems that in Fedora Core 4 they have
5664 set this on a number of new functions. No doubt there will be more in due
5667 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5668 specify a transport. However, if an address that was verified by such a
5669 router was the subject of a callout, Exim crashed because it tried to
5670 read the rcpt_include_affixes from the non-existent transport. Now it
5671 just assumes that the setting of that option is false. This bug was
5672 introduced by 4.51/PH/31.
5674 PH/28 Changed -d+all to exclude +memory, because that information is very
5675 rarely of interest, but it makes the output a lot bigger. People tend to
5676 do -d+all out of habit.
5678 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5679 code in os-type was giving problems when libc.so lives in lib64, like on
5682 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5683 aren't the modern standard, and it seems that some systems' include files
5684 don't always have them. Exim was already checking for some of the newer
5685 ones like T_AAAA, and defining it itself. I've added checks for all the
5686 record types that Exim uses.
5688 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5689 not automatically generating a new one, as it is supposed to. This
5690 prevented TLS from working. If the file did exist, but contained invalid
5691 data, a new version was generated, as expected. It was only the case of a
5692 non-existent file that was broken.
5694 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5695 with a change in libdomainkeys > 0.64.
5697 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5698 from DNS. If the selector record carries the flag, it now has
5699 precedence over the domain-wide flag.
5701 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5703 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5704 the use of a port name that isn't defined in /etc/services) occurred, the
5705 message was deferred as in a normal delivery, and thus remained on the
5706 spool, instead of being failed because of the mua_wrapper setting. This
5707 is now fixed, and I tidied up some of the mua_wrapper messages at the
5710 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5711 the messages in a hash of arrays rather than using individual hashes.
5712 This is a bit cleaner and results in dramatic memory savings, albeit
5713 at a slight CPU cost.
5715 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5716 as requested by Marc Sherman.
5718 SC/10 Eximstats: added histograms for user specified patterns as requested
5721 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5723 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5724 fopen() in the content-scanning modules that did not already have it.
5730 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5731 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5733 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5735 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5737 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5738 to test Sieve filters that use "vacation".
5740 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5741 that changes the way the GnuTLS parameters are stored in the cache file.
5742 The new format can be generated externally. For backward compatibility,
5743 if the data in the cache doesn't make sense, Exim assumes it has read an
5744 old-format file, and it generates new data and writes a new file. This
5745 means that you can't go back to an older release without removing the
5748 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5749 work if there are any delivery delays because "one_time" forces the
5750 parent to be marked "delivered", so its unseen clone is never tried
5751 again. For this reason, Exim now forbids the simultaneous setting of
5754 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5755 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5756 entry if you want to know the details. The fix, however, neglected to
5757 consider the case when local delivery batching is involved. The test for
5758 "previously delivered" was not happening when checking to see if an
5759 address could be batched with a previous (undelivered) one; under
5760 certain circumstances this could lead to multiple deliveries to the same
5763 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5764 in its include files, and this causes problems building Exim.
5766 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5767 header_syntax) but Exim was just ignoring anything given after a slash.
5768 In particular, this caused confusion with an attempt to use "verify =
5769 reverse_host_lookup/defer_ok". An error is now given when options are
5770 supplied for verify items that do not have them. (Maybe reverse_host_
5771 lookup should have a defer_ok option, but that's a different point.)
5773 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5774 defined by RFC 821) to 2048, because there were problems with some AUTH
5775 commands, and RFC 1869 says the size should be increased for extended
5776 SMTP commands that take arguments.
5778 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5781 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5782 "unknown" error; now it says that the functionality isn't in the binary.
5784 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5785 an address' error message when a string expansion fails (syntax or
5786 whatever). Otherwise the password may appear in the log. Following change
5787 PH/42 below, there is no longer a chance of it appearing in a bounce
5790 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5792 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5793 of its addresses was ever tried. (Bugzilla bug #2.)
5795 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5796 the result incorrectly in the debug output. (It correctly added a newline
5797 to what was transported.)
5799 TF/01 Added $received_time.
5801 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5802 commented out examples of how to interface to a virus scanner and to
5803 SpamAssassin. Also added commented examples of av_scanner and
5804 spamd_address settings.
5806 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5807 and controls are allowed in which ACLs. There were a couple of minor
5808 errors. Some of the entries in the conditions table (which is a table of
5809 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5810 negation of where the condition IS allowed.
5812 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5814 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5815 header file does not have a version number, so I've had to invent a new
5816 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5817 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5818 radiusclient), but it was contributed by a Radius user.
5820 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5821 files or directories, for OpenSSL.
5823 PH/19 When an Exim process that is running as root has to create an Exim log
5824 file, it does so in a subprocess that runs as exim:exim so as to get the
5825 ownership right at creation (otherwise, other Exim processes might see
5826 the file with the wrong ownership). There was no test for failure of this
5827 fork() call, which would lead to the process getting stuck as it waited
5828 for a non-existent subprocess. Forks do occasionally fail when resources
5829 run out. I reviewed all the other calls to fork(); they all seem to check
5832 PH/20 When checking for unexpected SMTP input at connect time (before writing
5833 the banner), Exim was not dealing correctly with a non-positive return
5834 from the read() function. If the client had disconnected by this time,
5835 the result was a log entry for a synchronization error with an empty
5836 string after "input=" when read() returned zero. If read() returned -1
5837 (an event I could not check), uninitialized data bytes were printed.
5838 There were reports of junk text (parts of files, etc) appearing after
5841 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5843 PH/22 Added support for macro redefinition, and (re)definition in between
5844 driver and ACL definitions.
5846 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5847 forgetting to use the resulting value; it was using the unexpanded value.
5849 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5850 hadn't been configured. The fix is from Juergen Kreileder, who
5851 understands it better than I do:
5853 "Here's what I see happening with three configured cyrus_sasl
5854 authenticators configured (plain, login, cram-md5):
5856 On startup auth_cyrus_sasl_init() gets called for each of these.
5857 This means three calls to sasl_listmech() without a specified mech_list.
5858 => SASL tests which mechs of all available mechs actually work
5859 => three warnings about OTP not working
5860 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5862 With the patch, sasl_listmech() also gets called three times. But now
5863 SASL's mech_list option is set to the server_mech specified in the the
5864 authenticator. Or in other words, the answer from sasl_listmech()
5865 gets limited to just the mech you're testing for (which is different
5867 => the return list contains just 'plain' or 'login', 'cram-md5' or
5868 nothing depending on the value of ob->server_mech.
5870 I've just tested the patch: Authentication still works fine,
5871 unavailable mechs specified in the exim configuration are still
5872 caught, and the auth.log warnings about OTP are gone."
5874 PH/25 When debugging is enabled, the contents of the command line are added
5875 to the debugging output, even when log_selector=+arguments is not
5878 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5879 answer is "GNU", and only if the return is "GNU/something" is the answer
5882 PH/27 $acl_verify_message is now set immediately after the failure of a
5883 verification in an ACL, and so is available in subsequent modifiers. In
5884 particular, the message can be preserved by coding like this:
5886 warn !verify = sender
5887 set acl_m0 = $acl_verify_message
5889 Previously, $acl_verify_message was set only while expanding "message"
5890 and "log_message" when a very denied access.
5892 PH/28 Modified OS/os.c-Linux with
5894 -#ifndef OS_LOAD_AVERAGE
5895 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5897 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5898 nomenclature these days.)
5900 PH/29 Installed patch from the Sieve maintainer that adds the options
5901 sieve_useraddress and sieve_subaddress to the redirect router.
5903 PH/30 In these circumstances:
5904 . Two addresses routed to the same list of hosts;
5905 . First host does not offer TLS;
5906 . First host accepts first address;
5907 . First host gives temporary error to second address;
5908 . Second host offers TLS and a TLS session is established;
5909 . Second host accepts second address.
5910 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5911 and peerdn, if requested) that were in fact used only for the second
5914 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5915 attention to any local part prefix or suffix that was matched by the
5916 router that accepted the address. It now behaves in the same way as it
5917 does for delivery: the affixes are removed from the local part unless
5918 rcpt_include_affixes is set on the transport.
5920 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5921 timeout during the DATA phase of an incoming message.
5923 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5924 applied a patch sanctioned by the Sieve maintainer.
5926 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5927 the uid or gid is negative. A case of a negative gid caused this to be
5928 noticed. The fix allows for either to be negative.
5930 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5931 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5934 PH/36 The change PH/12 above was broken. Fixed it.
5936 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5937 the grounds that routing the same address twice would always produce the
5938 same answer. This might have been true once, but it is certainly no
5939 longer true now. Routing a child address may depend on the previous
5940 routing that produced that child. Some complicated redirection strategies
5941 went wrong when messages had multiple recipients, and made Exim's
5942 behaviour dependent on the order in which the addresses were given.
5944 I have moved the duplicate checking until after the routing is complete.
5945 Exim scans the addresses that are assigned to local and remote
5946 transports, and removes any duplicates. This means that more work will be
5947 done, as duplicates will always all be routed, but duplicates are
5948 presumably rare, so I don't expect this is of any significance.
5950 For deliveries to pipes, files, and autoreplies, the duplicate checking
5951 still happens during the routing process, since they are not going to be
5954 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5955 It corrects a timeout issue with spamd. This is Ian's comment: "The
5956 background is that sometimes spamd either never reads data from a
5957 connection it has accepted, or it never writes response data. The exiscan
5958 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5959 blindly assumes that writes won't block so it may never time out."
5961 PH/39 Allow G after quota size as well as K and M.
5963 PH/40 The value set for $authenticated_id in an authenticator may not contain
5964 binary zeroes or newlines because the value is written to log lines and
5965 to spool files. There was no check on this. Now the value is run through
5966 the string_printing() function so that such characters are converted to
5967 printable escape sequences.
5969 PH/41 $message_linecount is a new variable that contains the total number of
5970 lines in the message. Compare $body_linecount, which is the count for the
5973 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5974 bounce and delay warning messages, except in certain special cases, which
5977 (a) An SMTP error message from a remote host;
5978 (b) A message specified in a :fail: redirection;
5979 (c) A message specified in a "fail" command in a system filter;
5980 (d) A message specified in a FAIL return from the queryprogram router;
5981 (e) A message specified by the cannot_route_message router option.
5983 In these cases only, Exim does include the error details in bounce and
5984 warning messages. There are also a few cases where bland messages such
5985 as "unrouteable address" or "local delivery error" are given.
5987 PH/43 $value is now also set for the "else" part of a ${run expansion.
5989 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5990 being worked on, but at least Exim now implements the latest version to
5993 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5994 process to complete was treated as a delivery failure, a timeout while
5995 writing the message to the pipe was logged, but erroneously treated as a
5996 successful delivery. Such timeouts include transport filter timeouts. For
5997 consistency with the overall process timeout, these timeouts are now
5998 treated as errors, giving rise to delivery failures by default. However,
5999 there is now a new Boolean option for the pipe transport called
6000 timeout_defer, which, if set TRUE, converts the failures into defers for
6001 both kinds of timeout. A transport filter timeout is now identified in
6004 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
6005 systems where "make" and "gmake" are different, calling "gmake" at top
6006 level broke things. I've arranged for the value of $(MAKE) to be passed
6007 from the Makefile to this script so that it can call the same version of
6011 A note about Exim versions 4.44 and 4.50
6012 ----------------------------------------
6014 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
6015 changes of various kinds. As a consequence, a big documentation update was
6016 needed. This delayed the release for rather longer than seemed good, especially
6017 in the light of a couple of (minor) security issues. Therefore, the changes
6018 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
6019 release. So 4.44 and 4.50 are in effect two different branches that both start
6022 I have left the 4.50 change log unchanged; it contains all the changes since
6023 4.43. The change log for 4.44 is below; many of its items are identical to
6024 those for 4.50. This seems to be the most sensible way to preserve the
6025 historical information.
6031 1. Minor wording change to the doc/README.SIEVE file.
6033 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
6034 computation of the current number of files was incorrect.
6036 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6037 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6040 4. Give more explanation in the error message when the command for a transport
6041 filter fails to execute.
6043 5. There are several places where Exim runs a non-Exim command in a
6044 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6045 was being done only for the command run by the queryprogram router. It is
6046 now done for all such subprocesses. The other cases are: ${run, transport
6047 filters, and the commands run by the lmtp and pipe transports.
6049 6. Added CONFIGURE_GROUP build-time option.
6051 7. Some older OS have a limit of 256 on the maximum number of file
6052 descriptors. Exim was using setrlimit() to set 1000 as a large value
6053 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6054 systems. I've change it so that if it can't get 1000, it tries for 256.
6056 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6057 was an oversight, and furthermore, ever since the addition of extra
6058 controls (e.g. 4.43/32), the checks on when to allow different forms of
6059 "control" were broken. There should now be diagnostics for all cases when a
6060 control that does not make sense is encountered.
6062 9. Added the /retain_sender option to "control=submission".
6064 10. $recipients is now available in the predata ACL (oversight).
6066 11. Tidy the search cache before the fork to do a delivery from a message
6067 received from the command line. Otherwise the child will trigger a lookup
6068 failure and thereby defer the delivery if it tries to use (for example) a
6069 cached ldap connection that the parent has called unbind on.
6071 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6072 of $address_data from the recipient verification was clobbered by the
6073 sender verification.
6075 13. The value of address_data from a sender verification is now available in
6076 $sender_address_data in subsequent conditions in the ACL statement.
6078 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6080 15. Added a new option "connect=<time>" to callout options, to set a different
6083 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6084 was its contents. (It was OK if the option was not defined at all.)
6086 17. A "Completed" log line is now written for messages that are removed from
6087 the spool by the -Mrm option.
6089 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6090 information about exactly what failed.
6092 19. Added -dd to debug only the daemon process.
6094 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6095 handles timeouts, both on the server side and network timeouts. Renamed the
6096 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6098 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6099 It is now set to "smtps".
6101 22. $host_address is now set to the target address during the checking of
6102 ignore_target_hosts.
6104 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6105 being passed; this would have caused $sender_host_name to have been used if
6106 matching the list had actually called for a host name (not very likely,
6107 since this list is usually IP addresses). A host name is now passed as
6110 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6111 code that specifies a non-restarting handler (typically sigaction() in
6112 modern systems) in an attempt to fix a rare and obscure crash bug.
6114 25. Narrowed the window for a race in the daemon that could cause it to ignore
6115 SIGCHLD signals. This is not a major problem, because they are used only to
6116 wake it up if nothing else does.
6118 26. A malformed maildirsize file could cause Exim to calculate negative values
6119 for the mailbox size or file count. Odd effects could occur as a result.
6120 The maildirsize information is now recalculated if the size or filecount
6123 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6124 support for a long time. Removed HAVE_SYS_VFS_H.
6126 28. Installed the latest version of exipick from John Jetmore.
6128 29. In an address list, if the pattern was not a regular expression, an empty
6129 subject address (from a bounce message) matched only if the pattern was an
6130 empty string. Non-empty patterns were not even tested. This was the wrong
6131 because it is perfectly reasonable to use an empty address as part of a
6132 database query. An empty address is now tested by patterns that are
6133 lookups. However, all the other forms of pattern expect the subject to
6134 contain a local part and a domain, and therefore, for them, an empty
6135 address still always fails if the pattern is not itself empty.
6137 30. Exim went into a mad DNS loop when attempting to do a callout where the
6138 host was specified on an smtp transport, and looking it up yielded more
6139 than one IP address.
6141 31. Re-factored the code for checking spool and log partition space into a
6142 function that finds that data and another that does the check. The former
6143 is then used to implement four new variables: $spool_space, $log_space,
6144 $spool_inodes, and $log_inodes.
6146 32. The RFC2047 encoding function was originally intended for short strings
6147 such as real names; it was not keeping to the 75-character limit for
6148 encoded words that the RFC imposes. It now respects the limit, and
6149 generates multiple encoded words if necessary. To be on the safe side, I
6150 have increased the buffer size for the ${rfc2047: expansion operator from
6153 33. It is now permitted to omit both strings after an "if" condition; if the
6154 condition is true, the result is "true". As before, when the second string
6155 is omitted, a false condition yields an empty string. This makes it less
6156 cumbersome to write custom ACL and router conditions.
6158 34. Failure to deliver a bounce message always caused it to be frozen, even if
6159 there was an errors_to setting on the router. The errors_to setting is now
6162 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6163 canonical form (fully expanded) before being placed in
6164 $sender_host_address.
6166 36. The table in the code that translates DNS record types into text (T_A to
6167 "A" for instance) was missing entries for NS and CNAME. It is just possible
6168 that this could have caused confusion if both these types were looked up
6169 for the same domain, because the text type is used as part of Exim's
6170 per-process caching. But the chance of anyone hitting this buglet seems
6173 37. The dnsdb lookup has been extended in a number of ways.
6175 (1) There is a new type, "zns", which walks up the domain tree until it
6176 finds some nameserver records. It should be used with care.
6178 (2) There is a new type, "mxh", which is like "mx" except that it returns
6179 just the host names, not the priorities.
6181 (3) It is now possible to give a list of domains (or IP addresses) to be
6182 looked up. The behaviour when one of the lookups defers can be
6183 controlled by a keyword.
6185 (4) It is now possible to specify the separator character for use when
6186 multiple records are returned.
6188 38. The dnslists ACL condition has been extended: it is now possible to supply
6189 a list of IP addresses and/or domains to be looked up in a particular DNS
6192 39. Added log_selector=+queue_time_overall.
6194 40. When running the queue in the test harness, wait just a tad after forking a
6195 delivery process, to get repeatability of debugging output.
6197 41. Include certificate and key file names in error message when GnuTLS fails
6198 to set them up, because the GnuTLS error message doesn't include the name
6199 of the failing file when there is a problem reading it.
6201 42. Allow both -bf and -bF in the same test run.
6203 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6205 44. The "Exiscan patch" is now merged into the mainline Exim source.
6207 45. Sometimes the final signoff response after QUIT could fail to get
6208 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6209 < 0 before doing a fflush(). This bug looks as though it goes back to the
6210 introduction of TLS in release 3.20, but "sometimes" must have been rare
6211 because the tests only now provoked it.
6213 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6214 (this can affect the format of dates).
6216 47. exim_tidydb, when checking for the continued existence of a message for
6217 which it has found a message-specific retry record, was not finding
6218 messages that were in split spool directories. Consequently, it was
6219 deleting retry records that should have stayed in existence.
6221 48. Steve fixed some bugs in eximstats.
6223 49. The SPA authentication driver was not abandoning authentication and moving
6224 on to the next authenticator when an expansion was forced to fail,
6225 contradicting the general specification for all authenticators. Instead it
6226 was generating a temporary error. It now behaves as specified.
6228 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6229 (the order specifies the preference for clients). The order is now AES256,
6230 AES128, 3DES, ARCFOUR128.
6232 51. Small patch to Sieve code - explicitly set From: when generating an
6235 52. Exim crashed if a remote delivery caused a very long error message to be
6236 recorded - for instance if somebody sent an entire SpamAssassin report back
6237 as a large number of 550 error lines. This bug was coincidentally fixed by
6238 increasing the size of one of Exim's internal buffers (big_buffer) that
6239 happened as part of the Exiscan merge. However, to be on the safe side, I
6240 have made the code more robust (and fixed the comments that describe what
6243 53. Now that there can be additional text after "Completed" in log lines (if
6244 the queue_time_overall log selector is set), a one-byte patch to exigrep
6245 was needed to allow it to recognize "Completed" as not the last thing in
6248 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6249 patch that reportedly fixes this has been added. I am not expert enough to
6250 create a test for it. This is what the patch creator wrote:
6252 "I found a little strange behaviour of ldap code when working with
6253 Windows 2003 AD Domain, where users was placed in more than one
6254 Organization Units. When I tried to give exim partial DN, the exit code
6255 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6256 But simultaneously result of request was absolutely normal ldap result,
6257 so I produce this patch..."
6259 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6260 REFERENCE, so I have modified the code to exclude the patch when that macro
6263 55. Some experimental protocols are using DNS PTR records for new purposes. The
6264 keys for these records are domain names, not reversed IP addresses. The
6265 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6266 leaves it alone. Component reversal etc. now happens only for IP addresses.
6269 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6271 57. Double the size of the debug message buffer (to 2048) so that more of very
6272 long debug lines gets shown.
6274 58. The exicyclog utility now does better if the number of log files to keep
6275 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6277 59. Two changes related to the smtp_active_hostname option:
6279 (1) $smtp_active_hostname is now available as a variable.
6280 (2) The default for smtp_banner uses $smtp_active_hostname instead
6281 of $primary_hostname.
6283 60. The host_aton() function is supposed to be passed a string that is known
6284 to be a valid IP address. However, in the case of IPv6 addresses, it was
6285 not checking this. This is a hostage to fortune. Exim now panics and dies
6286 if the condition is not met. A case was found where this could be provoked
6287 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6288 components; fortuitously, this particular loophole had already been fixed
6289 by change 4.50/55 above.
6291 If there are any other similar loopholes, the new check in host_aton()
6292 itself should stop them being exploited. The report I received stated that
6293 data on the command line could provoke the exploit when Exim was running as
6294 exim, but did not say which command line option was involved. All I could
6295 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6296 running as the user.
6299 61. There was a buffer overflow vulnerability in the SPA authentication code
6300 (which came originally from the Samba project). I have added a test to the
6301 spa_base64_to_bits() function which I hope fixes it.
6304 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6305 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6307 63. The daemon start-up calls getloadavg() while still root for those OS that
6308 need the first call to be done as root, but it missed one case: when
6309 deliver_queue_load_max is set with deliver_drop_privilege. This is
6310 necessary for the benefit of the queue runner, because there is no re-exec
6311 when deliver_drop_privilege is set.
6313 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6314 This has been fixed.
6316 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6317 was in use, was not putting the data itself into the right store pool;
6318 consequently, it could be overwritten for a subsequent message in the same
6319 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6322 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6324 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6325 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6326 the function were treating the return as a boolean value, which happened to
6327 work because 0=false and not-0=true, but is not correct code.
6329 68. The host_aton() function was not handling scoped IPv6 addresses (those
6330 with, for example, "%eth0" on the end) correctly.
6332 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6333 negated items (that is, ~something) in unsigned ints. Some compilers
6334 apparently mutter when there is no cast.
6336 70. If an address verification called from an ACL failed, and did not produce a
6337 user-specific message (i.e. there was only a "system" message), nothing was
6338 put in $acl_verify_message. In this situation, it now puts the system
6341 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6342 session; change 4.31/43 added the unwanted input to the log line - except
6343 that it did not do this in the start of session case. It now does.
6345 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6346 This is wrong and can cause the other end to generate a synchronization
6347 error if it is another Exim or anything else that does the synchronization
6348 check. A QUIT command is no longer sent after a timeout.
6350 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6351 during host lookups.
6353 74. The defer_ok option of callout verification was not working if it was used
6354 when verifying addresses in header lines, that is, for this case:
6356 verify = header_sender/callout=defer_ok
6358 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6359 those file descriptors could be used for SMTP connections. If anything
6360 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6361 could be sent to the SMTP client, causing chaos. The daemon now opens
6362 stdin, stdout, and stderr to /dev/null when it puts itself into the
6365 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6366 log by default. The user can override this with suitable Perl magic.
6368 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6369 the log message when discard triggers, was not working for the DATA ACL or
6370 for the non-SMTP ACL.
6372 78. Error message wording change in sieve.c.
6374 79. If smtp_accept_max_per_host was set, the number of connections could be
6375 restricted to fewer than expected, because the daemon was trying to set up
6376 a new connection before checking whether the processes handling previous
6377 connections had finished. The check for completed processes is now done
6378 earlier. On busy systems, this bug wouldn't be noticed because something
6379 else would have woken the daemon, and it would have reaped the completed
6382 80. If a message was submitted locally by a user whose login name contained one
6383 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6384 It caused a spool format error. I have fixed the spool reading code. A
6385 related problem was that the "from" clause in the Received: line became
6386 illegal because of the space(s). It is now covered by ${quote_local_part.
6388 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6391 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6393 83. Patch from Sieve maintainer to fix unterminated string problem in
6394 "vacation" handling.
6396 84. Some minor changes to the Linux configuration files to help with other
6397 OS variants using glibc.
6399 85. One more patch for Sieve to update vacation handling to latest spec.
6402 ----------------------------------------------------
6403 See the note above about the 4.44 and 4.50 releases.
6404 ----------------------------------------------------
6410 1. Change 4.43/35 introduced a bug that caused file counts to be
6411 incorrectly computed when quota_filecount was set in an appendfile
6414 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6415 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6418 3. Give more explanation in the error message when the command for a transport
6419 filter fails to execute.
6421 4. There are several places where Exim runs a non-Exim command in a
6422 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6423 was being done only for the command run by the queryprogram router. It is
6424 now done for all such subprocesses. The other cases are: ${run, transport
6425 filters, and the commands run by the lmtp and pipe transports.
6427 5. Some older OS have a limit of 256 on the maximum number of file
6428 descriptors. Exim was using setrlimit() to set 1000 as a large value
6429 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6430 systems. I've change it so that if it can't get 1000, it tries for 256.
6432 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6433 was an oversight, and furthermore, ever since the addition of extra
6434 controls (e.g. 4.43/32), the checks on when to allow different forms of
6435 "control" were broken. There should now be diagnostics for all cases when a
6436 control that does not make sense is encountered.
6438 7. $recipients is now available in the predata ACL (oversight).
6440 8. Tidy the search cache before the fork to do a delivery from a message
6441 received from the command line. Otherwise the child will trigger a lookup
6442 failure and thereby defer the delivery if it tries to use (for example) a
6443 cached ldap connection that the parent has called unbind on.
6445 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6446 of $address_data from the recipient verification was clobbered by the
6447 sender verification.
6449 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6450 was its contents. (It was OK if the option was not defined at all.)
6452 11. A "Completed" log line is now written for messages that are removed from
6453 the spool by the -Mrm option.
6455 12. $host_address is now set to the target address during the checking of
6456 ignore_target_hosts.
6458 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6459 being passed; this would have caused $sender_host_name to have been used if
6460 matching the list had actually called for a host name (not very likely,
6461 since this list is usually IP addresses). A host name is now passed as
6464 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6465 code that specifies a non-restarting handler (typically sigaction() in
6466 modern systems) in an attempt to fix a rare and obscure crash bug.
6468 15. Narrowed the window for a race in the daemon that could cause it to ignore
6469 SIGCHLD signals. This is not a major problem, because they are used only to
6470 wake it up if nothing else does.
6472 16. A malformed maildirsize file could cause Exim to calculate negative values
6473 for the mailbox size or file count. Odd effects could occur as a result.
6474 The maildirsize information is now recalculated if the size or filecount
6477 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6478 support for a long time. Removed HAVE_SYS_VFS_H.
6480 18. Updated exipick to current release from John Jetmore.
6482 19. Allow an empty sender to be matched against a lookup in an address list.
6483 Previously the only cases considered were a regular expression, or an
6486 20. Exim went into a mad DNS lookup loop when doing a callout where the
6487 host was specified on the transport, if the DNS lookup yielded more than
6490 21. The RFC2047 encoding function was originally intended for short strings
6491 such as real names; it was not keeping to the 75-character limit for
6492 encoded words that the RFC imposes. It now respects the limit, and
6493 generates multiple encoded words if necessary. To be on the safe side, I
6494 have increased the buffer size for the ${rfc2047: expansion operator from
6497 22. Failure to deliver a bounce message always caused it to be frozen, even if
6498 there was an errors_to setting on the router. The errors_to setting is now
6501 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6502 canonical form (fully expanded) before being placed in
6503 $sender_host_address.
6505 24. Updated eximstats to version 1.33
6507 25. Include certificate and key file names in error message when GnuTLS fails
6508 to set them up, because the GnuTLS error message doesn't include the name
6509 of the failing file when there is a problem reading it.
6511 26. Expand error message when OpenSSL has problems setting up cert/key files.
6514 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6515 (this can affect the format of dates).
6517 28. exim_tidydb, when checking for the continued existence of a message for
6518 which it has found a message-specific retry record, was not finding
6519 messages that were in split spool directories. Consequently, it was
6520 deleting retry records that should have stayed in existence.
6522 29. eximstats updated to version 1.35
6523 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6524 1.35 - bugfix such that pie charts by volume are generated correctly
6526 30. The SPA authentication driver was not abandoning authentication and moving
6527 on to the next authenticator when an expansion was forced to fail,
6528 contradicting the general specification for all authenticators. Instead it
6529 was generating a temporary error. It now behaves as specified.
6531 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6532 (the order specifies the preference for clients). The order is now AES256,
6533 AES128, 3DES, ARCFOUR128.
6535 31. Small patch to Sieve code - explicitly set From: when generating an
6538 32. Exim crashed if a remote delivery caused a very long error message to be
6539 recorded - for instance if somebody sent an entire SpamAssassin report back
6540 as a large number of 550 error lines. This bug was coincidentally fixed by
6541 increasing the size of one of Exim's internal buffers (big_buffer) that
6542 happened as part of the Exiscan merge. However, to be on the safe side, I
6543 have made the code more robust (and fixed the comments that describe what
6546 33. Some experimental protocols are using DNS PTR records for new purposes. The
6547 keys for these records are domain names, not reversed IP addresses. The
6548 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6549 leaves it alone. Component reversal etc. now happens only for IP addresses.
6552 34. The host_aton() function is supposed to be passed a string that is known
6553 to be a valid IP address. However, in the case of IPv6 addresses, it was
6554 not checking this. This is a hostage to fortune. Exim now panics and dies
6555 if the condition is not met. A case was found where this could be provoked
6556 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6557 components; fortuitously, this particular loophole had already been fixed
6558 by change 4.50/55 or 4.44/33 above.
6560 If there are any other similar loopholes, the new check in host_aton()
6561 itself should stop them being exploited. The report I received stated that
6562 data on the command line could provoke the exploit when Exim was running as
6563 exim, but did not say which command line option was involved. All I could
6564 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6565 running as the user.
6568 35. There was a buffer overflow vulnerability in the SPA authentication code
6569 (which came originally from the Samba project). I have added a test to the
6570 spa_base64_to_bits() function which I hope fixes it.
6573 36. The daemon start-up calls getloadavg() while still root for those OS that
6574 need the first call to be done as root, but it missed one case: when
6575 deliver_queue_load_max is set with deliver_drop_privilege. This is
6576 necessary for the benefit of the queue runner, because there is no re-exec
6577 when deliver_drop_privilege is set.
6579 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6580 was in use, was not putting the data itself into the right store pool;
6581 consequently, it could be overwritten for a subsequent message in the same
6582 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6585 38. Sometimes the final signoff response after QUIT could fail to get
6586 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6587 < 0 before doing a fflush(). This bug looks as though it goes back to the
6588 introduction of TLS in release 3.20, but "sometimes" must have been rare
6589 because the tests only now provoked it.
6595 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6596 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6597 Now it returns an integer. A number of calls were still expecting a T/F
6598 return. Fortuitously, in all cases, the tests worked in OK situations,
6599 which is the norm. However, things would have gone wrong on any write
6600 failures on the smtp file descriptor. This function is used when sending
6601 messages over SMTP and also when doing verify callouts.
6603 2. When Exim is called to do synchronous delivery of a locally submitted
6604 message (the -odf or -odi options), it no longer closes stderr before doing
6607 3. Implemented the mua_wrapper option.
6609 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6611 5. Implemented the functions header_remove(), header_testname(),
6612 header_add_at_position(), and receive_remove_recipient(), and exported them
6615 6. If an ACL "warn" statement specified the addition of headers, Exim already
6616 inserted X-ACL-Warn: at the start if there was no header name. However, it
6617 was not making this test for the second and subsequent header lines if
6618 there were newlines in the string. This meant that an invalid header could
6619 be inserted if Exim was badly configured.
6621 7. Allow an ACL "warn" statement to add header lines at the start or after all
6622 the Received: headers, as well as at the end.
6624 8. Added the rcpt_4xx retry error code.
6626 9. Added postmaster_mailfrom=xxx to callout verification option.
6628 10. Added mailfrom=xxxx to the callout verification option, for verify=
6631 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6632 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6634 12. Inserted some casts to stop certain compilers warning when using pointer
6635 differences as field lengths or precisions in printf-type calls (mostly
6636 affecting debugging statements).
6638 13. Added optional readline() support for -be (dynamically loaded).
6640 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6641 same clock tick as a message's arrival, so that its received time was the
6642 same as the "first fail" time on the retry record, and that message
6643 remained on the queue past the ultimate address timeout, every queue runner
6644 would try a delivery (because it was past the ultimate address timeout) but
6645 after another failure, the ultimate address timeout, which should have then
6646 bounced the address, did not kick in. This was a "< instead of <=" error;
6647 in most cases the first failure would have been in the next clock tick
6648 after the received time, and all would be well.
6650 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6651 being recognized when the domain list was tested by the match_domain
6652 condition in an expansion string.
6654 16. Added the ${str2b64: operator.
6656 17. Exim was always calling setrlimit() to set a large limit for the number of
6657 processes, without checking whether the existing limit was already
6658 adequate. (It did check for the limit on file descriptors.) Furthermore,
6659 errors from getrlimit() and setrlimit() were being ignored. Now they are
6660 logged to the main and panic logs, but Exim does carry on, to try to do its
6661 job under whatever limits there are.
6663 18. Imported PCRE 5.0.
6665 19. Trivial typo in log message " temporarily refused connection" (the leading
6668 20. If the log selector return_path_on_delivery was set and an address was
6669 redirected to /dev/null, the delivery process crashed because it assumed
6670 that a return path would always be set for a "successful" delivery. In this
6671 case, the whole delivery is bypassed as an optimization, and therefore no
6674 21. Internal re-arrangement: the function for sending a challenge and reading
6675 a response while authentication was assuming a zero-terminated challenge
6676 string. It's now changed to take a pointer and a length, to allow for
6677 binary data in such strings.
6679 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6681 23. Exim was not respecting finduser_retries when seeking the login of the
6682 uid under which it was called; it was always trying 10 times. (The default
6683 setting of finduser_retries is zero.) Also, it was sleeping after the final
6684 failure, which is pointless.
6686 24. Implemented tls_on_connect_ports.
6688 25. Implemented acl_smtp_predata.
6690 26. If the domain in control=submission is set empty, Exim assumes that the
6691 authenticated id is a complete email address when it generates From: or
6692 Sender: header lines.
6694 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6695 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6696 chown and chgrp in /bin and hostname in /usr/bin.
6698 28. Exim was keeping the "process log" file open after each use, just as it
6699 does for the main log. This opens the possibility of it remaining open for
6700 long periods when the USR1 signal hits a daemon. Occasional processlog
6701 errors were reported, that could have been caused by this. Anyway, it seems
6702 much more sensible not to leave this file open at all, so that is what now
6705 29. The long-running daemon process does not normally write to the log once it
6706 has entered its main loop, and it closes the log before doing so. This is
6707 so that log files can straightforwardly be renamed and moved. However,
6708 there are a couple of unusual error situations where the daemon does write
6709 log entries, and I had neglected to close the log afterwards.
6711 30. The text of an SMTP error response that was received during a remote
6712 delivery was being truncated at 512 bytes. This is too short for some of
6713 the long messages that one sometimes sees. I've increased the limit to
6716 31. It is now possible to make retry rules that apply only when a message has a
6717 specific sender, in particular, an empty sender.
6719 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6720 it possible to be selective about when SMTP synchronization is enforced.
6722 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6724 32. Implemented hosts_connection_nolog.
6726 33. Added an ACL for QUIT.
6728 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6731 35. Added mailbox_size and mailbox_filecount to appendfile.
6733 36. Added control = no_multiline_responses to ACLs.
6735 37. There was a bug in the logic of the code that waits for the clock to tick
6736 in the case where the clock went backwards by a substantial amount such
6737 that the microsecond fraction of "now" was more than the microsecond
6738 fraction of "then" (but the whole seconds number was less).
6740 38. Added support for the libradius Radius client library this is found on
6741 FreeBSD (previously only the radiusclient library was supported).
6747 1. When certain lookups returned multiple values in the form name=value, the
6748 quoting of the values was not always being done properly. Specifically:
6749 (a) If the value started with a double quote, but contained no whitespace,
6751 (b) If the value contained whitespace other than a space character (i.e.
6752 tabs or newlines or carriage returns) it was not quoted.
6753 This fix has been applied to the mysql and pgsql lookups by writing a
6754 separate quoting function and calling it from the lookup code. The fix
6755 should probably also be applied to nisplus, ibase and oracle lookups, but
6756 since I cannot test any of those, I have not disturbed their existing code.
6758 2. A hit in the callout cache for a specific address caused a log line with no
6759 reason for rejecting RCPT. Now it says "Previous (cached) callout
6760 verification failure".
6762 3. There was an off-by-one bug in the queryprogram router. An over-long
6763 return line was truncated at 256 instead of 255 characters, thereby
6764 overflowing its buffer with the terminating zero. As well as fixing this, I
6765 have increased the buffer size to 1024 (and made a note to document this).
6767 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6768 when Exim is waiting for an SMTP response from a remote server, Exim
6769 restarts its select() call on the socket, thereby resetting its timeout.
6770 This is not a problem when such interrupts are rare. Somebody set up a cron
6771 job to run exiwhat every 2 minutes, which is less than the normal select()
6772 timeout (5 or 10 minutes). This meant that the select() timeout never
6773 kicked in because it was always reset. I have fixed this by comparing the
6774 time when an interrupt arrives with the time at the start of the first call
6775 to select(). If more time than the timeout has elapsed, the interrupt is
6776 treated as a timeout.
6778 5. Some internal re-factoring in preparation for the addition of Sieve
6779 extensions (by MH). In particular, the "personal" test is moved to a
6780 separate function, and given an option for scanning Cc: and Bcc: (which is
6781 not set for Exim filters).
6783 6. When Exim created an email address using the login of the caller as the
6784 local part (e.g. when creating a From: or Sender: header line), it was not
6785 quoting the local part when it contained special characters such as @.
6787 7. Installed new OpenBSD configuration files.
6789 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6790 try to make them clearer.
6792 9. Callout options, other than the timeout value, were being ignored when
6793 verifying sender addresses in header lines. For example, when using
6795 verify = header_sender/callout=no_cache
6797 the cache was (incorrectly) being used.
6799 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6800 only the Cygwin environment.
6802 11. When return_path_on_delivery was set as a log selector, if different remote
6803 addresses in the same message used different return paths and parallel
6804 remote delivery occurred, the wrong values would sometimes be logged.
6805 (Whenever a remote delivery process finished, the return path value from
6806 the most recently started remote delivery process was logged.)
6808 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6809 lines when AUTH and/or TLS are in use. This is the "received protocol"
6810 field. Exim used to use "asmtp" for authenticated SMTP, without any
6811 indication (in the protocol name) for TLS use. Now it follows the RFC and
6812 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6813 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6814 names appear in log lines as well as in Received: header lines.
6816 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6817 extensions, and comparison tests, and to fix some bugs.
6819 14. Changes to the "personal" filter test:
6821 (1) The test was buggy in that it was just doing the equivalent of
6822 "contains" tests on header lines. For example, if a user's address was
6823 anne@some.where, the "personal" test would incorrectly be true for
6825 To: susanne@some.where
6827 This test is now done by extracting each address from the header in turn,
6828 and checking the entire address. Other tests that are part of "personal"
6829 are now done using regular expressions (for example, to check local parts
6830 of addresses in From: header lines).
6832 (2) The list of non-personal local parts in From: addresses has been
6833 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6834 taken from the Sieve specification recommendations.
6836 (3) If the message contains any header line starting with "List-" it is
6837 treated as non-personal.
6839 (4) The test for "circular" in the Subject: header line has been removed
6840 because it now seems ill-conceived.
6842 15. Minor typos in src/EDITME comments corrected.
6844 16. Installed latest exipick from John Jetmore.
6846 17. If headers_add on a router specified a text string that was too long for
6847 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6848 of string_sprintf() is now avoided.
6850 18. $message_body_size was not set (it was always zero) when running the DATA
6851 ACL and the local_scan() function.
6853 19. For the "mail" command in an Exim filter, no default was being set for
6854 the once_repeat time, causing a random time value to be used if "once" was
6855 specified. (If the value happened to be <= 0, no repeat happened.) The
6856 default is now 0s, meaning "never repeat". The "vacation" command was OK
6857 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6858 (I found it when inspecting the code).
6860 20. There is now an overall timeout for performing a callout verification. It
6861 defaults to 4 times the callout timeout, which applies to individual SMTP
6862 commands during the callout. The overall timeout applies when there is more
6863 than one host that can be tried. The timeout is checked before trying the
6864 next host. This prevents very long delays if there are a large number of
6865 hosts and all are timing out (e.g. when the network connections are timing
6866 out). The value of the overall timeout can be changed by specifying an
6867 additional sub-option for "callout", called "maxwait". For example:
6869 verify = sender/callout=5s,maxwait=20s
6871 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6872 to the end before writing, but this should make it even safer).
6874 22. Exim was forgetting that it had advertised PIPELINING for the second and
6875 subsequent messages on an SMTP connection. It was also not resetting its
6876 memory on STARTTLS and an internal HELO.
6878 23. When Exim logs an SMTP synchronization error within a session, it now
6879 records whether PIPELINING has been advertised or not.
6881 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6882 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6883 rather than long int.
6885 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6887 26. Added the never_mail option to autoreply.
6893 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6894 crash if the getsockname() call failed; this can happen if a connection is
6895 closed very soon after it is established. The problem was simply in the
6896 order in which certain operations were done, causing Exim to try to write
6897 to the SMTP stream before it had set up the file descriptor. The bug has
6898 been fixed by making things happen in the correct order.
6904 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6905 before the connection was closed, thus losing the rejection response.
6907 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6908 some early Solaris releases, but causes trouble in current releases where
6909 socklen_t is defined.
6911 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6914 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6917 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6918 that releasing the top of it at the end releases what was used for sub-
6919 expansions (unless the block got too big). However, discard this block if
6920 the first thing is a variable or header, so that we can use its block when
6921 it is dynamic (useful for very large $message_headers, for example).
6923 6. Lookups now cache *every* query, not just the most recent. A new, separate
6924 store pool is used for this. It can be recovered when all lookup caches are
6925 flushed. Lookups now release memory at the end of their result strings.
6926 This has involved some general refactoring of the lookup sources.
6928 7. Some code has been added to the store_xxx() functions to reduce the amount
6929 of flapping under certain conditions.
6931 8. log_incoming_interface used to affect only the <= reception log lines. Now
6932 it causes the local interface and port to be added to several more SMTP log
6933 lines, for example "SMTP connection from", and rejection lines.
6935 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6937 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6939 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6940 could be overwritten at the end of the current message (or the start of a
6941 new message if it was set in a HELO ACL). The value is now preserved for
6942 the duration of the SMTP connection.
6944 12. If a transport had a headers_rewrite setting, and a matching header line
6945 contained an unqualified address, that address was qualified, even if it
6946 did not match any rewriting rules. The underlying bug was that the values
6947 of the flags that permit the existence of unqualified sender and recipient
6948 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6949 non-local messages, and by -bnq for local messages) were not being
6950 preserved with the message after it was received.
6952 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6953 "next input=" as part of the text comprising the host identity instead of
6954 the correct text. The code was using the same buffer for two different
6955 strings. However, depending on which order the printing function evaluated
6956 its arguments, the bug did not always show up. Under Linux, for example, my
6957 test suite worked just fine.
6959 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6960 doesn't work with some older versions of Perl. It has been changed to "my",
6961 which in any case is probably the better facility to use.
6963 15. A really picky compiler found some instances of statements for creating
6964 error messages that either had too many or two few arguments for the format
6967 16. The size of the buffer for calls to the DNS resolver has been increased
6968 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6969 for addresses that have a lot of PTR records. This alleviates a problem; it
6970 does not fully solve it.
6972 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6973 buffer now truncates the list and logs the incident, which is the same
6974 action as happens when Exim is looking up a host name and its aliases.
6975 Previously in this situation something unpredictable would happen;
6976 sometimes it was "internal error: store_reset failed".
6978 18. If a server dropped the connection unexpectedly when an Exim client was
6979 using GnuTLS and trying to read a response, the client delivery process
6980 crashed while trying to generate an error log message.
6982 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6983 string, for example:
6985 warn message = H1: something\nH2: something
6987 the text was added as a single header line from Exim's point of view
6988 though it ended up OK in the delivered message. However, searching for the
6989 second and subsequent header lines using $h_h2: did not work. This has been
6990 fixed. Similarly, if a system filter added multiple headers in this way,
6991 the routers could not see them.
6993 20. Expanded the error message when iplsearch is called with an invalid key to
6994 suggest using net-iplsearch in a host list.
6996 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6997 ACLs are no longer actually imposed (and a message to that effect is
7000 22. If a "gecos" field in a passwd entry contained escaped characters, in
7001 particular, if it contained a \" sequence, Exim got it wrong when building
7002 a From: or a Sender: header from that name. A second bug also caused
7003 incorrect handling when an unquoted " was present following a character
7004 that needed quoting.
7006 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
7007 was not being matched caselessly.
7009 24. Arranged for all hyphens in the exim.8 source to be escaped with
7012 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
7013 properly. Recipient callout cache records were still being keyed to include
7014 the sender, even when use_sender was set false. This led to far more
7015 callouts that were necessary. The sender is no longer included in the key
7016 when use_sender is false.
7018 26. Added "control = submission" modifier to ACLs.
7020 27. Added the ${base62d: operator to decode base 62 numbers.
7022 28. dnsdb lookups can now access SRV records.
7024 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
7025 the configuration file.
7027 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
7028 (-v) mode. This makes the output for a verbose queue run more intelligible.
7030 31. Added a use_postmaster feature to recipient callouts.
7032 32. Added the $body_zerocount variable, containing the number of binary zero
7033 bytes in the message body.
7035 33. The time of last modification of the "new" subdirectory is now used as the
7036 "mailbox time last read" when there is a quota error for a maildir
7039 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
7041 35. Added +ignore_unknown as a special item in host lists.
7043 36. Code for decoding IPv6 addresses in host lists is now included, even if
7044 IPv6 support is not being compiled. This fixes a bug in which an IPv6
7045 address was recognized as an IP address, but was then not correctly decoded
7046 into binary, causing unexpected and incorrect effects when compared with
7053 1. Very minor rewording of debugging text in manualroute to say "list of
7054 hosts" instead of "hostlist".
7056 2. If verify=header_syntax was set, and a header line with an unqualified
7057 address (no domain) and a large number of spaces between the end of the
7058 name and the colon was received, the reception process suffered a buffer
7059 overflow, and (when I tested it) crashed. This was caused by some obsolete
7060 code that should have been removed. The fix is to remove it!
7062 3. When running in the test harness, delay a bit after writing a bounce
7063 message to get a bit more predictability in the log output.
7065 4. Added a call to search_tidyup() just before forking a reception process. In
7066 theory, someone could use a lookup in the expansion of smtp_accept_max_
7067 per_host which, without the tidyup, could leave open a database connection.
7069 5. Added the variables $recipient_data and $sender_data which get set from a
7070 lookup success in an ACL "recipients" or "senders" condition, or a router
7071 "senders" option, similar to $domain_data and $local_part_data.
7073 6. Moved the writing of debug_print from before to after the "senders" test
7076 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7077 problems for message scanning, either using a data ACL, or using
7078 local_scan() because the Received: header was not generated till after they
7079 were called (in order to set the time as the time of reception completion).
7080 I have revised the way this works. The header is now generated after the
7081 body is received, but before the ACL or local_scan() are called. After they
7082 are run, the timestamp in the header is updated.
7088 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7089 before starting a queue runner without re-exec. This happened only when
7090 deliver_drop_privilege was set or when the Exim user was set to root. The
7091 effect of the bug was that timeouts during subsequent deliveries caused
7092 crashes instead of being properly handled. The handler is now left at its
7093 default (and expected) setting.
7095 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7096 message, again when deliver_drop_privilege is set or Exim is run as root.
7097 The bug described in (1) was not present in this case, but the tidying up
7098 of the other signals was missing. I have made the two cases consistent.
7100 3. The ignore_target_hosts setting on a manualroute router was being ignored
7101 for hosts that were looked up using the /MX notation.
7103 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7106 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7107 operated on the sender address. After changing the $sender_address to <>
7108 for the sender address verify, Exim was re-instated it as the original
7109 (before rewriting) address, but remembering that it had rewritten it, so it
7110 wasn't rewriting it again. This bug also had the effect of breaking the
7111 sender address verification caching when the sender address was rewritten.
7113 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7114 This has been changed so that if the ip literal address matches
7115 ignore_target_hosts, the router declines.
7117 7. Added expansion conditions match_domain, match_address, and match_local_
7118 part (NOT match_host).
7120 8. The placeholder for the Received: header didn't have a length field set.
7122 9. Added code to Exim itself and to exim_lock to test for a specific race
7123 condition that could lead to file corruption when using MBX delivery. The
7124 issue is with the lockfile that is created in /tmp. If this file is removed
7125 after a process has opened it but before that process has acquired a lock,
7126 there is the potential for a second process to recreate the file and also
7127 acquire a lock. This could lead to two Exim processes writing to the file
7128 at the same time. The added code performs the same test as UW imapd; it
7129 checks after acquiring the lock that its file descriptor still refers to
7130 the same named file.
7132 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7133 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7136 11. Added the smtp_active_hostname option. If used, this will typically be made
7137 to depend on the incoming interface address. Because $interface_address is
7138 not set up until the daemon has forked a reception process, error responses
7139 that can happen earlier (such as "too many connections") no longer contain
7142 12. If an expansion in a condition on a "warn" statement fails because a lookup
7143 defers, the "warn" statement is abandoned, and the next ACL statement is
7144 processed. Previously this caused the whole ACL to be aborted.
7146 13. Added the iplsearch lookup type.
7148 14. Added ident_timeout as a log selector.
7150 15. Added tls_certificate_verified as a log selector.
7152 16. Added a global option tls_require_ciphers (compare the smtp transport
7153 option of the same name). This controls incoming TLS connections.
7155 17. I finally figured out how to make tls_require_ciphers do a similar thing
7156 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7157 before starting the TLS session.
7159 18. Tabs are now shown as \t in -bP output.
7161 19. If the log selector return_path_on_delivery was set, Exim crashed when
7162 bouncing a message because it had too many Received: header lines.
7164 20. If two routers both had headers_remove settings, and the first one included
7165 a superfluous trailing colon, the final name in the first list and the
7166 first name in the second list were incorrectly joined into one item (with a
7167 colon in the middle).
7173 1. Added -C and -D options to the exinext utility, mainly to make it easier
7174 to include in the automated testing, but these could be helpful when
7175 multiple configurations are in use.
7177 2. The exinext utility was not formatting the output nicely when there was
7178 an alternate port involved in the retry record key, nor when there was a
7179 message id as well (for retries that were specific to a specific message
7180 and a specific host). It was also confused by IPv6 addresses, because of
7181 the additional colons they contain. I have fixed the IPv4 problem, and
7182 patched it up to do a reasonable job for IPv6.
7184 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7185 delivery, the log line now contains "pipelined" if PIPELINING was used.
7187 4. An SMTP transport process used to panic and die if the bind() call to set
7188 an explicit outgoing interface failed. This has been changed; it is now
7189 treated in the same way as a connect() failure.
7191 5. A reference to $sender_host_name in the part of a conditional expansion
7192 that was being skipped was still causing a DNS lookup. This no longer
7195 6. The def: expansion condition was not recognizing references to header lines
7196 that used bh_ and bheader_.
7198 7. Added the _cache feature to named lists.
7200 8. The code for checking quota_filecount in the appendfile transport was
7201 allowing one more file than it should have been.
7203 9. For compatibility with Sendmail, the command line option
7211 and sets the incoming protocol and host name (for trusted callers). The
7212 host name and its colon can be omitted when only the protocol is to be set.
7213 Note the Exim already has two private options, -pd and -ps, that refer to
7214 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7215 "s", but I don't think that's a major issue.
7217 10. A number of refactoring changes to the code, none of which should affect
7220 (a) The number of logging options was getting close to filling up the
7221 32-bit word that was used as a bit map. I have split them into two classes:
7222 those that are passed in the argument to log_write(), and those that are
7223 only ever tested independently outside of that function. These are now in
7224 separate 32-bit words, so there is plenty of room for expansion again.
7225 There is no change in the user interface or the logging behaviour.
7227 (b) When building, for example, log lines, the code previously used a
7228 macro that called string_cat() twice, in order to add two strings. This is
7229 not really sufficiently general. Furthermore, there was one instance where
7230 it was actually wrong because one of the argument was used twice, and in
7231 one call a function was used. (As it happened, calling the function twice
7232 did not affect the overall behaviour.) The macro has been replaced by a
7233 function that can join an arbitrary number of extra strings onto a growing
7236 (c) The code for expansion conditions now uses a table and a binary chop
7237 instead of a serial search (which was left over from when there were very
7238 few conditions). Also, it now recognizes conditions like "pam" even when
7239 the relevant support is not compiled in: a suitably worded error message is
7240 given if an attempt is made to use such a condition.
7242 11. Added ${time_interval:xxxxx}.
7244 12. A bug was causing one of the ddress fields not to be passed back correctly
7245 from remote delivery subprocesses. The field in question was not being
7246 subsequently used, so this caused to problems in practice.
7248 13. Added new log selectors queue_time and deliver_time.
7250 14. Might have fixed a bug in maildirsizefile handling that threw up
7251 "unexpected character" debug warnings, and recalculated the data
7252 unnecessarily. In any case, I expanded the warning message to give more
7255 15. Added the message "Restricted characters in address" to the statements in
7256 the default ACL that block characters like @ and % in local parts.
7258 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7259 Three changes have been made:
7261 (a) There was a serious bug; a negative response to MAIL caused the whole
7262 recipient domain to be cached as invalid, thereby blocking all messages
7263 to all local parts at the same domain, from all senders. This bug has
7264 been fixed. The domain is no longer cached after a negative response to
7265 MAIL if the sender used is not empty.
7267 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7270 (c) A new callout option, "use_sender" has been added for people who want
7271 the modified behaviour.
7277 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7280 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7281 indeed breaks things for older releases.
7283 3. Added additional logging to the case where there is a problem reading data
7284 from a filter that is running in a subprocess using a pipe, in order to
7285 try to track down a specific problem.
7287 4. Testing facility fudge: when running in the test harness and attempting
7288 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7289 getting "No route to host". Convert this to a timeout.
7291 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7294 6. Some OS don't have socklen_t but use size_t instead. This affects the
7295 fifth argument of getsockopt() amongst other things. This is now
7296 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7297 can be set for individual OS. I have set it for SunOS5, OSF1, and
7298 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7299 some earlier ones do not.
7301 7. Change 4.30/15 was not doing the test caselessly.
7303 8. The standard form for an IPv6 address literal was being rejected by address
7304 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7305 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7306 this, as well as the form without the "IPv6" on the front (but only when
7307 address literals are enabled, of course).
7309 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7311 10. Exim crashed if a message with an empty sender address specified by -f
7312 encountered a router with an errors_to setting. This could be provoked only
7313 by a command such as
7317 where an empty string was supplied; "<>" did not hit this bug.
7319 11. Installed PCRE release 4.5.
7321 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7322 remained set. It is now erased.
7324 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7325 times from message ids (which are base 36 rather than the normal 62).
7327 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7328 were being counted as actual protocol errors, and logged if the log
7329 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7330 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7331 and DATA following a set of rejected RCPTs do not count as protocol errors.
7332 In other words, Exim assumes they were pipelined, though this may not
7333 actually be the case. Of course, in all cases the client gets an
7334 appropriate error code.
7336 15. If a lookup fails in an ACL condition, a message about the failure may
7337 be available; it is used if testing the ACL cannot continue, because most
7338 such messages specify what the cause of the deferral is. However, some
7339 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7340 that caused an old message to be retained and used if a later statement
7341 caused a defer, replacing the real cause of the deferral.
7343 16. If an IP address had so many PTR records that the DNS lookup buffer
7344 was not large enough to hold them, Exim could crash while trying to process
7345 the truncated data. It now detects and logs this case.
7347 17. Further to 4.21/58, another change has been made: if (and only if) the
7348 first line of a message (the first header line) ends with CRLF, a bare LF
7349 in a subsequent header line has a space inserted after it, so as not to
7350 terminate the header.
7352 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7353 unnecessarily, used atoi() instead of strtol(), and didn't check the
7354 termination when getting file sizes from file names by regex.
7356 19. Completely re-implemented the support for maildirsize files, in the light
7357 of a number of problems with the previous contributed implementation
7358 (4.30/29). In particular:
7360 . If the quota is zero, the maildirsize file is maintained, but no quota is
7363 . If the maildir directory does not exist, it is created before any attempt
7364 to write a maildirsize file.
7366 . The quota value in the file is just a cache; if the quota is changed in
7367 the transport, the new value overrides.
7369 . A regular expression is available for excluding directories from the
7372 20. The autoreply transport checks the characters in options that define the
7373 message's headers; it allows continued headers, but it was checking with
7374 isspace() after an embedded newline instead of explicitly looking for a
7377 21. If all the "regular" hosts to which an address was routed had passed their
7378 expiry times, and had not reached their retry times, the address was
7379 bounced, even if fallback hosts were defined. Now Exim should go on to try
7382 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7383 equivalent code in the SMTP transport. Some hosts send humungous responses
7384 to HELO/EHLO, more than 1024 it seems.
7386 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7387 gives compiler warnings in some environments. I've now done it "properly",
7390 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7391 (because of problems with the built-in one) was declared to return uschar *
7392 instead of char *, causing compiler failure.
7394 25. Fixed a file descriptor leak when processing alias/forward files.
7396 26. Fixed a minor format string issue in dbfn.c.
7398 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7400 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7401 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7404 29. When a sender address is verified, it is cached, to save repeating the test
7405 when there is more than one recipient in a message. However, when the
7406 verification involves a callout, it is possible for different callout
7407 options to be set for different recipients. It is too complicated to keep
7408 track of this in the cache, so now Exim always runs a verification when a
7409 callout is required, relying on the callout cache for the optimization.
7410 The overhead is duplication of the address routing, but this should not be
7413 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7414 to be verified with callout=postmaster, and the main callout worked but the
7415 postmaster check failed, the verification correctly failed. However, if a
7416 subsequent RCPT command asked for sender verification *without* the
7417 postmaster check, incorrect caching caused this verification also to fail,
7420 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7421 it was not caching the DNS options (qualify_single, search_parents) that
7422 were used when the lookup failed. A subsequent lookup with different
7423 options therefore always gave the same answer, though there were cases
7424 where it should not have. (Example: a "domains = !$mx_any" option on a
7425 dnslookup router: the "domains" option is always processed without any
7426 widening, but the router might have qualify_single set.) Now Exim uses the
7427 cached value only when the same options are set.
7429 32. Added John Jetmore's "exipick" utility to the distribution.
7431 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7432 than a timeout (e.g. a certificate is required, and is not provided), an
7433 Exim server now closes the connection immediately. Previously it waited for
7434 the client to close - but if the client is SSL, it seems that they each
7435 wait for each other, leading to a delay before one of them times out.
7437 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7438 maintained 0.8.x compatibility because I don't think many are using it, and
7439 it is clearly obsolete.
7441 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7444 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7445 client certificate was expired. A simple patch fixes this, though I don't
7446 understand the full logic of why the verify callback is called multiple
7449 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7450 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7451 which causes problems with some clients (such as the Certicom SSL Plus
7452 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7453 disables the coutermeasure allowing Eudora to connect."
7455 38. Exim was not checking that a write() to a log file succeeded. This could
7456 lead to Bad Things if a log got too big, in particular if it hit a file
7457 size limit. Exim now panics and dies if it cannot write to a log file, just
7458 as it does if it cannot open a log file.
7460 39. Modified OS/Makefile-Linux so that it now contains
7462 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7464 The two -D definitions ensure that Exim is compiled with large file
7465 support, which makes it possible to handle log files that are bigger than
7468 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7469 instance) a domain was checked against a named list that involved a lookup,
7470 causing $domain_data to be set, then another domain was checked against the
7471 same list, then the first domain was re-checked, the value of $domain_data
7472 after the final check could be wrong. In particular, if the second check
7473 failed, it could be set empty. This bug probably also applied to
7476 41. The strip_trailing_dot option was not being applied to the address given
7477 with the -f command-line option.
7479 42. The code for reading a message's header from the spool was incrementing
7480 $received_count, but never initializing it. This meant that the value was
7481 incorrect (doubled) while delivering a message in the same process in which
7482 it was received. In the most common configuration of Exim, this never
7483 happens - a fresh exec is done - but it can happen when
7484 deliver_drop_privilege is set.
7486 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7487 it now includes up to 150 characters of the unexpected data in the log
7490 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7491 and building data strings. The size of both of these buffers was 10 000
7492 bytes - far larger than anybody would *ever* want, thought I. Needless to
7493 say, somebody hit the limit. I have increased the maximum line length to
7494 20 000 and the maximum data length of concatenated lines to 100 000. I have
7495 also fixed two bugs, because there was no checking on these buffers. Tsk,
7496 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7497 buffer is too small.
7499 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7500 lsearch lookups. Now it does.
7502 46. When parsing a route_list item in a manualroute router, a fixed-length
7503 buffer was used for the list of hosts. I made this 1024 bytes long,
7504 thinking that nobody would ever have a list of hosts that long. Wrong.
7505 Somebody had a whole pile of complicated expansion conditions, and the
7506 string was silently truncated, leading to an expansion error. It turns out
7507 that it is easier to change to an unlimited length (owing to other changes
7508 that have happened since this code was originally written) than to build
7509 structure for giving a limitation error. The length of the item that
7510 expands into the list of hosts is now unlimited.
7512 47. The lsearch lookup could not handle data where the length of text line was
7513 more than 4095 characters. Such lines were truncated, leading to shortened
7514 data being returned. It should now handle lines of any length.
7516 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7517 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7520 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7521 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7522 no longer happens in comment lines. A list of replacements is now placed
7523 at the head of all of the source files, except those whose only change is
7524 to replace PERL_COMMAND in the very first #! line.
7526 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7527 messages on the queue, with a bottom-up merge sort, using code contributed
7528 by Michael Haardt. This should make operations like -bp somewhat faster on
7529 large queues. It won't affect queue runners, except when queue_run_in_order
7532 51. Installed eximstats 1.31 in the distribution.
7534 52. Added support for SRV lookups to the dnslookup router.
7536 53. If an ACL referred to $message_body or $message_body_end, the value was not
7537 reset for any messages that followed in the same SMTP session.
7539 54. The store-handling optimization for building very long strings was not
7540 differentiating between the different store pools. I don't think this
7541 actually made any difference in practice, but I've tidied it.
7543 55. While running the routers to verify a sender address, $sender_address
7544 was still set to the sender address. This is wrong, because when routing to
7545 send a bounce to the sender, it would be empty. Therefore, I have changed
7546 it so that, while verifying a sender address, $sender_address is set to <>.
7547 (There is no change to what happens when verifying a recipient address.)
7549 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7550 target A or AAAA records (if not already returned) without resetting the
7551 qualify_single or search_parents options of the DNS resolver. These are
7552 inappropriate in this case because the targets of MX and SRV records must
7553 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7554 target that, when qualified, matched something in the local domain. These
7555 two options are now turned off when doing these lookups.
7557 57. It seems that at least some releases of Reiserfs (which does not have the
7558 concept of a fixed number of inodes) returns zero and not -1 for the
7559 number of available inodes. This interacted badly with check_spool_inodes,
7560 which assumed that -1 was the "no such thing" setting. What I have done is
7561 to check that the total number of inodes is greater than zero before doing
7562 the test of how many are available.
7564 58. When a "warn" ACL statement has a log_message modifier, the message is
7565 remembered, and not repeated. This is to avoid a lot of repetition when a
7566 message has many recipients that cause the same warning to be written.
7567 However, Exim was preserving the list of already written lines for an
7568 entire SMTP session, which doesn't seem right. The memory is now reset if a
7569 new message is started.
7571 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7572 debugging output unless log_rewrite was also set.
7574 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7575 of (int)(handle) when we know that handle contains (void *)(-1).
7577 61. The Exim daemon panic-logs an error return when it closes the incoming
7578 connection. However "connection reset by peer" seems to be common, and
7579 isn't really an error worthy of noting specially, so that particular error
7582 62. When Exim is trying to find all the local interfaces, it used to panic and
7583 die if the ioctl to get the interface flags failed. However, it seems that
7584 on at least one OS (Solaris 9) it is possible to have an interface that is
7585 included in the list of interfaces, but for which you get a failure error
7586 for this call. This happens when the interface is not "plumbed" into a
7587 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7588 failure of the "get flags" call assumes that the interface is down.
7590 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7591 makes life easier for people who are doing arithmetic on fields extracted
7592 from dates, where you often get leading zeros that should not be
7593 interpreted as octal.
7595 64. Added qualify_domain to the redirect router, to override the global
7598 65. If a pathologically long header line contained very many addresses (the
7599 report of this problem mentioned 10 000) and each of them was rewritten,
7600 Exim could use up a very large amount of memory. (It kept on making new
7601 copies of the header line as it rewrote, and never released the old ones.)
7602 At the expense of a bit more processing, the header rewriting function has
7603 been changed so that it no longer eats memory in this way.
7605 66. The generation of the Received: header has been moved from the time that a
7606 message starts to be received, to the time that it finishes. The timestamp
7607 in the Received: header should now be very close to that of the <= log
7608 line. There are two side-effects of this change:
7610 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7611 logged header lines no longer include the local Received: line, because
7612 it has not yet been created. The same applies to a copy of the message
7613 that is returned to a non-SMTP sender when a message is rejected.
7615 (b) When a filter file is tested using -bf, no additional Received: header
7616 is added to the test message. After some thought, I decided that this
7619 This change does not affect the value of $received_for. It is still set
7620 after address rewriting, but before local_scan() is called.
7622 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7624 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7625 gave an unhelpful panic error message, and a defer error. I have managed to
7626 change this behaviour so that it now rejects any supplied certificate,
7627 which seems right, as the list of acceptable certificates is empty.
7629 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7630 gave an unhelpful defer error. I have not managed to make this reject any
7631 supplied certificates, but the error message it gives is "no certificate
7632 supplied", which is not helpful.
7634 70. exigrep's output now also includes lines that are not associated with any
7635 message, but which match the given pattern. Implemented by a patch from
7636 Martin Sluka, which also tidied up the Perl a bit.
7638 71. Recipient callout verification, like sender verification, was using <> in
7639 the MAIL FROM command. This isn't really the right thing, since the actual
7640 sender may affect whether the remote host accepts the recipient or not. I
7641 have changed it to use the actual sender in the callout; this means that
7642 the cache record is now keyed on a recipient/sender pair, not just the
7643 recipient address. There doesn't seem to be a real danger of callout loops,
7644 since a callout by the remote host to check the sender would use <>.
7645 [SEE ABOVE: changed after hitting problems.]
7647 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7648 temporary errors. However, in the case of such a code being given after
7649 the end of a data transmission (i.e. after ".") Exim was failing to write
7650 a retry record for the message. (Yes, there was some broken host that was
7651 actually sending 8xx at this point.)
7653 73. An unknown lookup type in a host list could cause Exim to panic-die when
7654 the list was checked. (An example that provoked this was putting <; in the
7655 middle of a list instead of at the start.) If this happened during a DATA
7656 ACL check, a -D file could be left lying around. This kind of configuration
7657 error no longer causes Exim to die; instead it causes a defer error. The
7658 incident is still logged to the main and panic logs.
7660 74. Buglet left over from Exim 3 conversion. The message "too many messages
7661 in one connection" was written to the rejectlog but not the mainlog, except
7662 when address rewriting (yes!) was being logged.
7664 75. Added write_rejectlog option.
7666 76. When a system filter was run not as root (that is, when system_filter_user
7667 was set), the values of the $n variables were not being returned to the
7668 main process; thus, they were not subsequently available in the $sn
7671 77. Added +return_path_on_delivery log selector.
7673 78. A connection timeout was being treated differently from recipients deferred
7674 when testing hosts_max_try with a message that was older than the host's
7675 retry timeout. (The host should not be counted, thus allowing all hosts to
7676 be tried at least once before bouncing.) This may have been the cause of an
7677 occasionally reported bug whereby a message would remain on the queue
7678 longer than the retry timeout, but would be bounced if a delivery was
7679 forced. I say "may" because I never totally pinned down the problem;
7680 setting up timeout/retry tests is difficult. See also the next item.
7682 79. The ultimate address timeout was not being applied to errors that involved
7683 a combination of host plus message (for example, a timeout on a MAIL
7684 command). When an address resolved to a number of possible hosts, and they
7685 were not all tried for each delivery (e.g. because of hosts_max_try), a
7686 message could remain on the queue longer than the retry timeout.
7688 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7691 81. Fixed an obscure SMTP outgoing bug which required at least the following
7692 conditions: (a) there was another message waiting for the same server;
7693 (b) the server returned 5xx to all RCPT commands in the first message so
7694 that the message was not completed; (c) the server dropped the connection
7695 or gave a negative response to the RSET that Exim sends to abort the
7696 transaction. The observed case was a dropped connection after DATA that had
7697 been sent in pipelining mode. That is, the server had advertised PIPELINING
7698 but was not implementing it correctly. The effect of the bug was incorrect
7699 behaviour, such as trying another host, and this could lead to a crash.
7705 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7706 and daemon.c were passed as pointers to ints; they should have been
7707 pointers to socklen_t variables (which are typically unsigned ints).
7709 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7712 3. Fixed a really odd bug that affected only the testing scheme; patching a
7713 certain fixed string in the binary changed the value of another string that
7714 happened to be identical to the end of the original first string.
7716 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7717 name", it returns that address as the IP address. On some operating
7718 systems (e.g. Solaris), it also passes back the IP address string as the
7719 "host name". However, on others (e.g. Linux), it passes back an empty
7720 string. Exim wasn't checking for this, and was changing the host name to an
7721 empty string, assuming it had been canonicalized.
7723 5. Although rare, it is permitted to have more than one PTR record for a given
7724 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7725 all the names associated with an address, because they do in Solaris.
7726 However, it seems that they do not in Linux for data that comes from the
7727 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7728 I found this out when I moved to a new Linux workstation and tried to run
7729 the Exim test suite.
7731 To get round this problem I have changed the code so that it now does its
7732 own call to the DNS to look up PTR records when searching for a host name.
7733 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7734 addresses that are only in /etc/hosts are still found.
7736 This behaviour is, however, controlled by an option called host_lookup_
7737 order, which defaults to "bydns:byaddr". If people want to use the other
7738 order, or indeed, just use one or the other means of lookup, they can
7739 specify it in this variable.
7741 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7742 some operating systems, this comes back from gethostbyaddr() as an empty
7743 string, and this is what Exim used to test for. However, it seems that in
7744 other systems, "." is yielded. Exim now tests for this case too.
7746 7. The values of check_spool_space and check_log_space are now held internally
7747 as a number of kilobytes instead of an absolute number of bytes. If a
7748 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7749 kilobyte. This means that much larger values can be stored.
7751 8. Exim monitor: an attempt to get the action menu when not actually pointing
7752 at a message produces an empty menu entitled "No message selected". This
7753 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7754 no entries in it ("Shell widget menu has zero width and/or height"). So I
7755 have added a single, blank menu entry in this case.
7757 9. Added ${quote_local_part.
7759 10. MIME decoding is now applied to the contents of Subject: header lines when
7762 11. Now that a reference to $sender_host_address automatically causes a reverse
7763 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7764 host lookup before query-style lookups in lists that might use this
7765 variable. This has therefore been abolished, and the "net-" prefix is no
7766 longer necessary for query-style lookups.
7768 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7769 have been a typo for LFLAGS, so it has been changed.
7771 13. The install script calls Exim with "-C /dev/null" in order to find the
7772 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7773 to be output. However, since Exim outputs its version number before the
7774 error, it didn't break the script. It just looked ugly. I fixed this by
7775 always allowing "-C /dev/null" if the caller is root.
7777 14. Ignore overlarge ACL variable number when reading spool file - insurance
7778 against a later release with more variables having written the file.
7780 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7781 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7782 this, as well as the form without the "IPv6" on the front.
7784 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7785 OS/Makefile-Darwin file.
7787 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7788 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7790 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7791 variable. However, some OS use ints for inodes. I've added cast to long int
7792 to get rid of the compiler warning.
7794 19. I had forgotten to lock out "/../" in configuration file names when
7795 ALT_CONFIG_PREFIX was set.
7797 20. Routers used for verification do not need to specify transports. However,
7798 if such a router generated a host list, and callout was configured, Exim
7799 crashed, because it could not find a port number from the (non-existent)
7800 transport. It now assumes port 25 in this circumstance.
7802 21. Added the -t option to exigrep.
7804 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7805 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7806 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7807 them is set, LOOKUP_LSEARCH is forced.
7809 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7811 24. Added sender and host information to the "rejected by local_scan()" log
7812 line; previously there was no indication of these.
7814 25. Added .include_if_exists.
7816 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7817 Linux. It turns out that not all file systems support this. Apparently some
7818 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7819 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7820 not supported on the file descriptor, is now ignored when Exim is trying to
7821 sync a directory. This applies only to Linux.
7823 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7825 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7826 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7829 29. Added support for maildirsize files from supplied patch (modified a bit).
7831 30. The use of :fail: followed by an empty string could lead Exim to respond to
7832 sender verification failures with (e.g.):
7834 550 Verification failed for <xxx>
7835 550 Sender verify failed
7837 where the first response line was missing the '-' that indicates it is not
7838 the final line of the response.
7840 31. The loop for finding the name of the user that called Exim had a hardwired
7841 limit of 10; it now uses the value of finduser_retries, which is used for
7842 all other user lookups.
7844 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7847 33. Exim was neglecting to zero errno before one call of strtol() when
7848 expanding a string and expecting an integer value. On some systems this
7849 resulted in spurious "integer overflow" errors. Also, it was casting the
7850 result into an int without checking.
7852 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7853 did not work. The code looks as if it has *never* worked, though it appears
7854 to have been documented since at least release 1.62. I have made it work.
7856 35. The "timeout_DNS" error in retry rules, also documented since at least
7857 1.62, also never worked. As it isn't clear exactly what this means, and
7858 clearly it isn't a major issue, I have abolished the feature by treating it
7859 as "timeout", and writing a warning to the main and panic logs.
7861 36. The display of retry rules for -brt wasn't always showing the error code
7864 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7865 timeout_connect_A, timeout_connect_MX.
7867 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7868 to the empty sender.
7870 39. The daemon was not analysing the content of -oX till after it had closed
7871 stderr and disconnected from the controlling terminal. This meant that any
7872 syntax errors were only noted on the panic log, and the return code from
7873 the command was 0. By re-arranging the code a little, I've made the
7874 decoding happen first, so such errors now appear on stderr, and the return
7875 code is 1. However, the actual setting up of the sockets still happens in
7876 the disconnected process, so errors there are still only recorded on the
7879 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7880 connections (as happens on some IP stacks) was logged at start up time as
7881 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7882 it from "IPv6 and IPv4", which means that two separate sockets are being
7885 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7886 says whether AF_INET or AF_INET6 was passed as an argument.
7888 42. Exiwhat output was messed up when time zones were included in log
7891 43. Exiwhat now gives more information about the daemon's listening ports,
7892 and whether -tls-on-connect was used.
7894 44. The "port" option of the smtp transport is now expanded.
7896 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7897 silently ignored. Now an error message is written to the main and panic
7900 46. There's a new ACL modifier called "logwrite" which writes to a log file
7901 as soon as it is encountered.
7903 47. Added $local_user_uid and $local_user_gid at routing time.
7905 48. Exim crashed when trying to verify a sender address that was being
7908 49. Exim was recognizing only a space character after ".include". It now also
7909 recognizes a tab character.
7911 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7912 extracting the relevant information from the specification. The man page no
7913 longer contains scrambled data for the -d option, and I've added a section
7914 at the front about calling Exim under different names.
7916 51. Added "extra_headers" argument to the "mail" command in filter files.
7918 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7921 53. Installed eximstats 1.29.
7923 54. Added transport_filter_timeout as a generic transport option.
7925 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7926 Cc: header lines. This was required by RFC 822, but it not required by RFC
7929 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7930 incoming messages that did not have them. Now it does so only if the
7931 message originates locally, that is, if there is no associated remote host
7932 address. When Resent- header lines are present, this applies to the Resent-
7933 lines rather than the non-Resent- lines.
7935 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7936 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7938 58. Messages for configuration errors now include the name of the main
7939 configuration file - useful now that there may be more than one file in a
7940 list (.included file names were always shown).
7942 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7943 for those rare installations that do not start the daemon as root or run it
7944 setuid root. I've cut out the call to initgroups() if the daemon is not
7947 60. The Exim user and group can now be bound into the binary as text strings
7948 that are looked up at the start of Exim's processing.
7950 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7952 62. Added $mailstore_basename variable.
7954 63. Installed patch to sieve.c from Michael Haardt.
7956 64. When Exim failed to open the panic log after failing to open the main log,
7957 the original message it was trying to log was written to stderr and debug
7958 output, but if they were not available (the usual case in production), it
7959 was lost. Now it is written to syslog before the two lines that record the
7960 failures to open the logs.
7962 65. Users' Exim filters run in subprocesses under the user's uid. It is
7963 possible for a "deliver" command or an alias in a "personal" command to
7964 provoke an address rewrite. If logging of address rewriting is configured,
7965 this fails because the process is not running as root or exim. There may be
7966 a better way of dealing with this, but for the moment (because 4.30 needs
7967 to be released), I have disabled address rewrite logging when running a
7968 filter in a non-root, non-exim process.
7974 1. The buildconfig auxiliary program wasn't quoting the value set for
7975 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7976 not defined. This bug was masked in 4.22 by the effect that was fixed in
7979 2. Some messages that were rejected after a message id was allocated were
7980 shown as "incomplete" by exigrep. It no longer does this for messages that
7981 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7983 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7984 have allow_domain_literals set, the ID did not get logged in the <= line.
7985 Domain literals are now always recognized in Message-ID: header lines.
7987 4. The first argument for a ${extract expansion item is the key name or field
7988 number. Leading and trailing spaces in this item were not being ignored,
7989 causing some misleading effects.
7991 5. When deliver_drop_privilege was set, single queue runner processes started
7992 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7993 same command in the process it spins off) were not dropping privilege.
7995 6. When the daemon running as "exim" started a queue runner, it always
7996 re-executed Exim in the spun-off process. This is a waste of effort when
7997 deliver_drop_privilege is set. The new process now just calls the
7998 queue-runner function directly.
8004 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
8007 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
8008 ignored. Though the use of -oP was forcing the writing of a pid file, it
8009 was always written to the default place.
8011 3. If the message "no IP address found for host xxxx" is generated during
8012 incoming verification, it is now followed by identification of the incoming
8013 connection (so you can more easily find what provoked it).
8015 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
8017 5. Added some features to "harden" Exim a bit more against certain attacks:
8019 (a) There is now a build-time option called FIXED_NEVER_USERS that can
8020 be put in Local/Makefile. This is like the never_users runtime option,
8021 but it cannot be overridden. The default setting is "root".
8023 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
8024 prefix string with which any file named in a -C command line option
8027 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
8028 is retained for -C and -D only if the caller of Exim is root. Without
8029 it, the exim user may also use -C and -D and retain privilege.
8031 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
8032 command line option is disabled.
8034 6. Macro names set by the -D option must start with an upper case letter, just
8035 like macro names defined in the configuration file.
8037 7. Added "dereference=" facility to LDAP.
8039 8. Two instances of the typo "uknown" in the source files are fixed.
8041 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
8042 the Configure-Makefile script screwed up while processing it.
8044 10. Incorporated PCRE 4.4.
8046 11. The SMTP synchronization check was not operating right at the start of an
8047 SMTP session. For example, it could not catch a HELO sent before the client
8048 waited for the greeting. There is now a check for outstanding input at the
8049 point when the greeting is written. Because of the duplex, asynchronous
8050 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8051 way, but not yet received, when the check is performed.
8053 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8054 on TCP/IP sockets, because this apparently causes some broken clients to
8057 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8058 unchanged) from the Cygwin maintainer.
8060 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8061 was supported instead of testing for mbx. Effectively a typo.
8063 15. The spa authenticator server code was not checking that the input it
8064 received was valid base64.
8066 16. The debug output line for the "set" modifier in ACLs was not showing the
8067 name of the variable that was being set.
8069 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8071 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8072 lookup on its own. Something else was needed to trigger the lookup. For
8073 example, a match in host_lookup or the need for a host name in a host list.
8074 Now, if $sender_host_name is referenced and the host name has not yet been
8075 looked up, a lookup is performed. If the lookup fails, the variable remains
8076 empty, and $host_lookup_failed is set to "1".
8078 19. Added "eqi" as a case-independent comparison operator.
8080 20. The saslauthd authentication condition could segfault if neither service
8081 nor realm was specified.
8083 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8084 error message that was logged was misleading, and incoming SMTP
8085 connections were dropped. The message is now more accurate, and temporary
8086 errors are given to SMTP connections.
8088 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8089 (except RSET) with a 421 error, until QUIT is received. However, it was
8090 failing to send a response to QUIT.
8092 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8093 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8094 ACL is now run after the helo_try_verify_hosts code.
8096 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8097 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8098 case-independent, so other case variants are also recognized). Apparently
8099 some systems use these upper case variants.
8101 25. If more than two messages were waiting for the same host, and a transport
8102 filter was specified for the transport, Exim sent two messages over the
8103 same TCP/IP connection, and then failed with "socket operation on non-
8104 socket" when it tried to send the third.
8106 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8108 27. The extern definition of crypt16() in expand.c was not being excluded when
8109 the OS had its own crypt16() function.
8111 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8112 as a preferred synonym for return_size_limit, both as an option and as an
8115 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8117 30. Changed the default configuration ACL to relax the local part checking rule
8118 for addresses that are not in any local domains. For these addresses,
8119 slashes and pipe symbols are allowed within local parts, but the sequence
8120 /../ is explicitly forbidden.
8122 31. SPA server authentication was not clearing the challenge buffer before
8125 32. log_message in a "warn" ACL statement was writing to the reject log as
8126 well as to the main log, which contradicts the documentation and doesn't
8127 seem right (because no rejection is happening). So I have stopped it.
8129 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8130 However, I am unable to do any testing of this.
8132 34. Fixed an infelicity in the appendfile transport. When checking directories
8133 for a mailbox, to see if any needed to be created, it was accidentally
8134 using path names with one or more superfluous leading slashes; tracing
8135 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8137 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8138 contents are added to the log line that is written for every discarded
8139 recipient. (Previously a log_message setting was ignored.)
8141 36. The ${quote: operator now quotes the string if it is empty.
8143 37. The install script runs exim in order to find its version number. If for
8144 some reason other than non-existence or emptiness, which it checks, it
8145 could not run './exim', it was installing it with an empty version number,
8146 i.e. as "exim-". This error state is now caught, and the installation is
8149 38. An argument was missing from the function that creates an error message
8150 when Exim fails to connect to the socket for saslauthd authentication.
8151 This could cause Exim to crash, or give a corrupted message.
8153 39. Added isip, isip4, and isip6 to ${if conditions.
8155 40. The ACL variables $acl_xx are now saved with the message, and can be
8156 accessed later in routers, transports, and filters.
8158 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8159 strings in the file are not string-expanded.
8161 42. If a MAIL command specified a SIZE value that was too large to fit into an
8162 int variable, the check against message_size_limit failed. Such values are
8163 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8164 day this will have to be increased, but I don't think I want to be around
8165 when emails are that large.
8172 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8173 iconv() is not standard in FreeBSD.
8175 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8176 IPv6 enabled. The observed symptom was a segmentation fault on return from
8177 the function os_common_find_running_interfaces() in src/os.c.
8179 3. In the check_special_case() function in daemon.c I had used "errno" as an
8180 argument name, which causes warnings on some systems. This was basically a
8181 typo, since it was named "eno" in the comments!
8183 4. The code that waits for the clock to tick (at a resolution of some fraction
8184 of a second) so as to ensure message-id uniqueness was always waiting for
8185 at least one whole tick, when it could have waited for less. [This is
8186 almost certainly not relevant at current processor speeds, where it is
8187 unlikely to ever wait at all. But we try to future-proof.]
8189 5. The function that sleeps for a time interval that includes fractions of a
8190 second contained a race. It did not block SIGALRM between setting the
8191 timer, and suspending (a couple of lines later). If the interval was short
8192 and the sigsuspend() was delayed until after it had expired, the suspension
8193 never ended. On busy systems this could lead to processes getting stuck for
8196 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8197 process, before it forks any delivery processes. The open lookup caching
8198 mechanism meant that the open file or database connection was passed into
8199 the delivery process. The problem was that delivery processes always tidy
8200 up cached lookup data. This could cause a problem for the next delivery
8201 process started by the queue runner, because the external queue runner
8202 process does not know about the closure. So the next delivery process
8203 still has data in the lookup cache. In the case of a file lookup, there was
8204 no problem because closing a file descriptor in a subprocess doesn't affect
8205 the parent. However, if the lookup was caching a connection to a database,
8206 the connection was closed, and the second delivery process was likely to
8207 see errors such as "PGSQL: query failed: server closed the connection
8208 unexpectedly". The problem has been fixed by closing all cached lookups
8209 in a queue runner before running a delivery process.
8211 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8212 seem to have the "const" qualifier which it has on other OS. I've
8215 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8216 *of the same type* (client or server) with the same public name that an
8217 error should be diagnosed.
8219 9. When Exim looked up a host name for an IP address, but failed to find the
8220 original IP address when looking up the host name (a safety check), it
8221 output the message "<ip address> does not match any IP for NULL", which was
8222 confusing, to say the least. The bug was that the host name should have
8223 appeared instead of "NULL".
8225 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8226 user that is built into the binary, and the -C or -D options is used, root
8227 privilege is dropped before the configuration file is read. In addition,
8228 logging is switched to stderr instead of the normal log files. If the
8229 configuration then re-defines the Exim user, the unprivileged environment
8230 is probably not what is expected, so Exim logs a panic warning message (but
8233 However, if deliver_drop_privilege is set, the unprivileged state may well
8234 be exactly what is intended, so the warning has been cut out in that case,
8235 and Exim is allowed to try to write to its normal log files.
8241 1. smtp_return_error_details was not giving details for temporary sender
8242 or receiver verification errors.
8244 2. Diagnose a configuration error if two authenticators have the same public
8247 3. Exim used not to create the message log file for a message until the first
8248 delivery attempt. This could be confusing when incoming messages were held
8249 for policy or load reasons. The message log file is now created at the time
8250 the message is received, and an initial "Received" line is written to it.
8252 4. The automatically generated man page for command line options had a minor
8253 bug that caused no ill effects; however, a more serious problem was that
8254 the procedure for building the man page automatically didn't always
8255 operate. Consequently, release 4.20 contains an out-of-date version. This
8256 shouldn't happen again.
8258 5. When building Exim with embedded Perl support, the script that builds the
8259 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8260 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8262 6. The freeze_tell option was not being used for messages that were frozen on
8263 arrival, either by an ACL or by local_scan().
8265 7. Added the smtp_incomplete_transaction log selector.
8267 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8268 was accepting AUTH without a new EHLO.
8270 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8271 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8274 10. Logging of TCP/IP connections (when configured) now happens in the main
8275 daemon process instead of the child process, so that the TCP/IP connection
8276 count is more accurate (but it can never be perfect).
8278 11. The use of "drop" in a nested ACL was not being handled correctly in the
8279 outer ACL. Now, if condition failure induced by the nested "drop" causes
8280 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8281 or "require"), the connection is dropped.
8283 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8284 that yield "discard" can now be used with an "accept" or a "discard" verb,
8285 but an error is generated for any others (because I can't see a useful way
8286 to define what should happen).
8288 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8289 are now processed in the same way as lines in the Exim configuration file.
8290 In particular, continuation lines are supported.
8292 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8294 15. Added -ti meaning -t -i.
8296 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8297 domains, and warn by logging if others are found.
8299 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8300 returned from GIFCONF when Exim is trying to find the list of interfaces on
8301 a host. The code in os.c has been modified to copy each ifreq to an aligned
8302 structure in all cases.
8304 Also, in some cases, the returned ifreq's were being copied to a 'struct
8305 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8306 means the last couple of bytes of an IPv6 address could be chopped if the
8307 ifreq contained only a normal sockaddr (14 bytes storage).
8309 18. Named domain lists were not supported in the hosts_treat_as_local option.
8310 An entry such as +xxxx was not recognized, and was treated as a literal
8313 19. Ensure that header lines added by a DATA ACL are included in the reject log
8314 if the ACL subsequently rejects the message.
8316 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8317 MD5 (which is deprecated).
8319 21. When testing a filter file using -bf, Exim was writing a message when it
8320 took the sender from a "From " line in the message, but it was not doing so
8321 when it took $return_path from a Return-Path: header line. It now does.
8323 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8324 with a valid header line field name (a series of printing characters
8325 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8327 23. Changed "disc" in the source to "disk" to conform to the documentation and
8328 the book and for uniformity.
8330 24. Ignore Sendmail's -Ooption=value command line item.
8332 25. When execve() failed while trying to run a command in a pipe transport,
8333 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8334 could be confused with a return value of 69 from the command itself. This
8335 has been changed to 127, the value the shell returns if it is asked to run
8336 a non-existent command. The wording for the related log line suggests a
8337 non-existent command as the problem.
8339 26. If received_header_text expands to an empty string, do not add a Received:
8340 header line to the message. (Well, it adds a token one on the spool, but
8341 marks it "old" so that it doesn't get used or transmitted.)
8343 27. Installed eximstats 1.28 (addition of -nt option).
8345 28. There was no check for failure on the call to getsockname() in the daemon
8346 code. This can fail if there is a shortage of resources on the system, with
8347 ENOMEM, for example. A temporary error is now given on failure.
8349 29. Contrary to the C standard, it seems that in some environments, the
8350 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8351 program. Exim now does this explicitly; it affects the formatting of
8352 timestamps using strftime().
8354 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8355 complaints. I've now initialized all the variables, to avoid this.
8357 32. Header lines added by a system filter were not being "seen" during
8358 transport-time rewrites.
8360 33. The info_callback() function passed to OpenSSL is set up with type void
8361 (*)(SSL *, int, int), as described somewhere. However, when calling the
8362 function (actually a macro) that sets it up, the type void(*)() is
8363 expected. I've put in a cast to prevent warnings from picky compilers.
8365 34. If a DNS black list lookup found a CNAME record, but there were no A
8366 records associated with the domain it pointed at, Exim crashed.
8368 35. If a DNS black list lookup returned more than one A record, Exim ignored
8369 all but the first. It now scans all returned addresses if a particular IP
8370 value is being sought. In this situation, the contents of the
8371 $dnslist_value variable are a list of all the addresses, separated by a
8374 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8375 to accept a host name and all its aliases if the forward lookup for any of
8376 them yielded the IP address of the incoming connection. Now it accepts only
8377 those names whose forward lookup yields the correct IP address. Any other
8378 names are discarded. This closes a loophole whereby a rogue DNS
8379 administrator could create reverse DNS records to break through a
8380 wildcarded host restriction in an ACL.
8382 37. If a user filter or a system filter that ran in a subprocess used any of
8383 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8384 the wrong values were passed to the pipe command ($thisaddress had the
8385 value of $0, $0 had the value of $1, etc). This bug was introduced by
8386 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8388 38. Improved the line breaking for long SMTP error messages from ACLs.
8389 Previously, if there was no break point between 40 and 75 characters, Exim
8390 left the rest of the message alone. Two changes have been made: (a) I've
8391 reduced the minimum length to 35 characters; (b) if it can't find a break
8392 point between 35 and 75 characters, it looks ahead and uses the first one
8393 that it finds. This may give the occasional overlong line, but at least the
8394 remaining text gets split now.
8396 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8397 file descriptors might be low, and that setting 1000 would always raise it.
8398 It turns out that in some environments, the limit is already over 1000 and
8399 that lowering it causes trouble. So now Exim takes care not to decrease it.
8401 40. When delivering a message, the value of $return_path is set to $sender_
8402 address at the start of routing (routers may change the value). By an
8403 oversight, this default was not being set up when an address was tested by
8404 -bt or -bv, which affected the outcome if any router or filter referred to
8407 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8408 log only when "message" or "log_message" are set. However, if one of the
8409 conditions was an address verification, or a call to a nested ACL, the
8410 messages generated by the underlying test were being passed through. This
8411 no longer happens. The underlying message is available in $acl_verify_
8412 message for both "message" and "log_message" expansions, so it can be
8413 passed through if needed.
8415 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8416 new expansion $bh_ to give the encoded byte string without charset
8417 translation. Translation happens only if iconv() is available; HAVE_ICONV
8418 indicates this at build time. HEADERS_CHARSET gives the charset to
8419 translate to; headers_charset can change it in the configuration, and
8420 "headers charset" can change it in an individual filter file.
8422 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8423 that creates RFC 2047 encoded "words" labels them as that charset instead
8424 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8425 expansion operator; (ii) when Exim creates a From: line for a local
8426 message; (iii) when a header line is rewritten to include a "phrase" part.
8428 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8429 buggy, causing it to skip the first lines of messages whose message ID
8430 ended in 'D'. This would not have bitten before Exim release 4.14, because
8431 message IDs were unlikely to end in 'D' before then. The effect was to have
8432 incorrect size information for certain domains.
8434 45. #include "config.h" was missing at the start of the crypt16.c module. This
8435 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8438 46. If there was a timeout during a "random" callout check, Exim treated it as
8439 a failure of the random address, and carried on sending RSET and the real
8440 address. If the delay was just some slowness somewhere, the response to the
8441 original RCPT would be taken as a response to RSET and so on, causing
8442 mayhem of various kinds.
8444 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8445 when I implemented it. It didn't allow for the fact that some option values
8446 may legitimately be negative (e.g. size_addition), and it didn't even do
8447 the right test for positive values.
8449 48. Domain names in DNS records are case-independent. Exim always looks them up
8450 in lower case. Some resolvers return domain names in exactly the case they
8451 appear in the zone file, that is, they may contain uppercase letters. Not
8452 all resolvers do this - some return always lower case. Exim was treating a
8453 change of case by a resolver as a change of domain, similar to a widening
8454 of a domain abbreviation. This triggered its re-routing code and so it was
8455 trying to route what was effectively the same domain again. This normally
8456 caused routing to fail (because the router wouldn't handle the domain
8457 twice). Now Exim checks for this case specially, and just changes the
8458 casing of the domain that it ultimately uses when it transmits the message
8461 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8464 50. If a filter generated a file delivery with a non-absolute name (possible if
8465 no home directory exists for the router), the forbid_file option was not
8468 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8469 the existing equality matching.
8471 52. Exim was using ints instead of ino_t variables in some places where it was
8472 dealing with inode numbers.
8474 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8475 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8476 called TMPDIR, and if it finds it is different, it changes its value.
8478 54. The smtp_printf() function is now made available to local_scan() so
8479 additional output lines can be written before returning. There is also an
8480 smtp_fflush() function to enable the detection of a dropped connection.
8481 The variables smtp_input and smtp_batched_input are exported to
8484 55. Changed the default runtime configuration: the message "Unknown user"
8485 has been removed from the ACL, and instead placed on the localuser router,
8486 using the cannot_route_message feature. This means that any verification
8487 failures that generate their own messages won't get overridden. Similarly,
8488 the "Unrouteable address" message that was in the ACL for unverifiable
8489 relay addresses has also been removed.
8491 56. Added hosts_avoid_esmtp to the smtp transport.
8493 57. The exicyclog script was not checking for the esoteric option
8494 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8495 will work only if exicyclog is run under the appropriate euid.
8497 58. Following a discussion on the list, the rules by which Exim recognises line
8498 endings on incoming messages have been changed. The -dropcr and drop_cr
8499 options are now no-ops, retained only for backwards compatibility. The
8500 following line terminators are recognized: LF CRLF CR. However, special
8501 processing applies to CR:
8503 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8504 nor a local message in the state where . is a terminator.
8506 (ii) If a bare CR is encountered in a header line, an extra space is added
8507 after the line terminator so as not to end the header. The reasoning
8508 behind this is that bare CRs in header lines are most likely either
8509 to be mistakes, or people trying to play silly games.
8511 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8512 was being incorrectly given as 18 bytes larger than it should have been.
8513 This is a VOB (very old bug).
8515 60. This may never have affected anything current, but just in case it has:
8516 When the local host is found other than at the start of a list of hosts,
8517 the local host, those with the same MX, and any that follow, are discarded.
8518 When the list in question was part of a longer list of hosts, the following
8519 hosts (not currently being processed) were also being discarded. This no
8520 longer happens. I'm not sure if this situation could ever has previously
8523 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8526 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8527 header. This is something that is recommended in a new Internet Draft, and
8528 is something that is documented as being done by Sendmail. There are two
8529 possible values. For messages generated by the autoreply transport, Exim
8532 Auto-Submitted: auto-replied
8534 whereas for all other generated messages (e.g. bounces) it adds
8536 Auto-Submitted: auto-generated
8538 63. The "personal" condition in filters now includes a test for the
8539 Auto-Submitted: header. If it contains the string "auto-" the message it
8540 not considered personal.
8542 64. Added rcpt_include_affixes as a generic transport option.
8544 65. Added queue_only_override (default true).
8546 66. Added the syslog_duplication option.
8548 67. If what should have been the first header line of a message consisted of
8549 a space followed by a colon, Exim was mis-interpreting it as a header line.
8550 It isn't of course - it is syntactically invalid and should therefore be
8551 treated as the start of the message body. The misbehaviour could have
8552 caused a number of strange effects, including loss of data in subsequent
8553 header lines, and spool format errors.
8555 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8556 client host had authenticated. This control can now be exercised by an ACL
8557 for more flexibility.
8559 69. By default, callouts do not happen when testing with -bh. There is now a
8560 variant, -bhc, which does actually run the callout code, including
8561 consulting and updating the callout cache.
8563 70. Added support for saslauthd authentication, courtesy of Alexander
8566 71. If statvfs() failed on the spool or log directories while checking their
8567 size for availability, Exim confusingly gave the error "space shortage".
8568 Furthermore, in debugging mode it crashed with a floating point exception.
8569 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8570 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8571 serious problem, Exim now writes to the main and panic logs when this
8572 happens, with details of the failure. It then refuses to accept the
8573 incoming message, giving the message "spool directory problem" or "log
8574 directory problem" with a 421 code for SMTP messages.
8576 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8577 0, 1, and 2 exist, because some OS complain for execs without them (see
8578 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8579 descriptors. However, the code omitted to check that the open succeeded,
8580 causing mysterious errors if for some reason the permissions on /dev/null
8581 got screwed. Now Exim writes a message to the main and panic logs, and
8582 bombs out if it can't open /dev/null.
8584 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8585 interact so that it is all more flexible. It is supposed to remain
8586 backwards compatible. Also added extra_local_interfaces.
8588 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8589 to bomb out with an assertion failure - to the client this appears as a
8590 connection drop. This problem occurs in the part of the code that was taken
8591 from the Samba project. Fortunately, the assertion is in a very simple
8592 function, so I have fixed this by reproducing the function inline in the
8593 one place where it is called, and arranging for authentication to fail
8594 instead of killing the process with assert().
8596 75. The SPA client code was not working when the server requested OEM rather
8597 than Unicode encoding.
8599 76. Added code to make require_files with a specific uid setting more usable in
8600 the case where statting the file as root fails - usually a non-root-mounted
8601 NFS file system. When this happens and the failure is EACCES, Exim now
8602 forks a subprocess and does the per-uid checking as the relevant uid.
8604 77. Added process_log_path.
8606 78. If log_file_path was not explicitly set, a setting of check_log_space or
8607 check_log_inodes was ignored.
8609 79. If a space check for the spool or log partitions fails, the incident is now
8610 logged. Of course, in the latter case the data may get lost...
8612 80. Added the %p formatting code to string_format() so that it can be used to
8613 print addresses in debug_print(). Adjusted all the address printing in the
8614 debugging in store.c to use %p rather than %d.
8616 81. There was a concern that a line of code in smtp_in.c could overflow a
8617 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8618 initially expressed, the concern was not well-founded, because trailing
8619 spaces are removed early. However, if the trailing spaces were followed by
8620 a NULL, they did not get removed, so the overflow was possible. Two fixes
8623 (a) I re-wrote the offending code in a cleaner fashion.
8624 (b) If an incoming SMTP command contains a NULL character, it is rejected
8627 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8628 runs initgroups(), so that if the Exim user is in any additional groups,
8629 they will be used during message reception.
8635 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob