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 trtying 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 broking this, giving only
84 a 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 vriable
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.
100 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
101 SMTP connection" log lines.
103 JH/02 Option default value updates:
104 - queue_fast_ramp (main) true (was false)
105 - remote_max_parallel (main) 4 (was 2)
107 JH/03 Cache static regex pattern compilations, for use by ACLs.
109 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
110 Make the rewrite never match and keep the logging. Trust the
111 admin to be using verify=header-syntax (to actually reject the message).
113 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
114 (under Linux) we watch the dir containing the final file; previously
115 it would be the dir with the first symlink. We still do not monitor
118 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
119 dn_expand() is happy to pass them through.
121 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
122 the file with the proof had an unchanged name, the new proof(s) were
123 loaded on top of the old ones (and nover used; the old ones were stapled).
125 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
126 more than one message arrived in a single connection a reference from
127 the earlier message could be re-used. Often a sigsegv resulted.
128 These variables were introduced in Exim 4.87.
129 Debug help from Graeme Fowler.
131 JH/09 Fix ${filter } for conditions that modify $value. Previously the
132 modified version would be used in construction the result, and a memory
135 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
136 Find and fix by Jasen Betts.
138 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
139 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
140 the systemwide configuration to override the Exim config.
142 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
143 API changes in libopendmarc.
145 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
146 pid 1, in the normal "background daemon" mode, having to drop process-
147 group leadership also lost track of needing to create listener sockets.
149 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
150 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
152 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
153 a capture group which obtained no text (eg. "(abc)*" matching zero
154 occurrences) could cause a segfault if the corresponding $<n> was
157 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
158 included a close-brace character (eg. it itself used an expansion) an
161 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
162 starting TLS. Previously it was after, meaning that attackers on such
163 ports had to be screened using the host_reject_connection main config
164 option. The new sequence aligns better with the STARTTLS behaviour, and
165 permits defences against crypto-processing load attacks, even though it
166 is strictly an incompatible change.
167 Also, avoid sending any SMTP fail response for either the connect ACL
168 or host_reject_connection, for TLS-on-connect ports.
170 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
171 Previously this was not permitted, but it makes reasonable sense.
172 While there, restore a restriction on using it from a connect ACL; given
173 the change JH/16 it could only return false (and before 4.91 was not
176 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
177 was exactly sized compared to the log buffer, a crash occurred with the
178 misleading message "bad memory reference; pool not found".
179 Found and traced by Jasen Betts.
181 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
182 dns_again_means_nonexist included an element causing a DNS lookup which
183 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
184 included (though probably not limited to) a process crash from stack
185 memory limit, or from excessive open files. Replace this with a paniclog
186 whine (as this is likely a configuration error), and returning
189 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
190 this always failed, probably leading to the usual downgrade to in-clear
193 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
194 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
195 to not checking that list for these lookups.
197 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
200 JH/23 Fix crash in string expansions. Previously, if an empty variable was
201 immediately followed by an expansion operator, a null-indirection read
202 was done, killing the process.
204 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
205 include an SMTP response string which is longer than that supported
206 by the delivering transport. Alleviate by wrapping such lines before
209 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
210 chars (RFC limit). Previously a limit of 12 items was made, which with
211 a not-impossible References: in the message being bounced could still
212 be over-large and get stopped in the transport.
214 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
215 close. Previously a bare socket close was done.
217 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
220 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
221 message being considered for delivery was already being handled by
222 another process, and having an SMTP connection already open, the function
223 to close it tried to use an uninitialized variable. This would afftect
224 high-volume sites more, especially when running mailing-list-style loads.
225 Pollution of logs was the major effect, as the other process delivered
226 the message. Found and partly investigated by Graeme Fowler.
228 JH/29 Change format of the internal ID used for message identification. The old
229 version only supported 31 bits for a PID element; the new 64 (on systems
230 which can use Base-62 encoding, which is all currently supported ones
231 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
232 and must use Base-36). The new ID is 23 characters rather than 16, and is
233 visible in various places - notably logs, message headers, and spool file
234 names. Various of the ancillary utilities also have to know the format.
235 As well as the expanded PID portion, the sub-second part of the time
236 recorded in the ID is expanded to support finer precision. Theoretically
237 this permits a receive rate from a single comms channel of better than the
239 The major timestamp part of the ID is not changed; at 6 characters it is
240 usable until about year 3700.
241 Updating from previously releases is fully supported: old-format spool
242 files are still usable, and the utilities support both formats. New
243 message will use the new format. The one hints-DB file type which uses
244 message-IDs (the transport wait- DB) will be discarded if an old-format ID
245 is seen; new ones will be built with only new-format IDs.
246 Optionally, a utility can be used to convert spool files from old to new,
247 but this is only an efficiency measure not a requirement for operation
248 Downgrading from new to old requires running a provided utility, having
249 first stopped all operations. This will convert any spool files from new
250 back to old (losing time-precision and PID information) and remove any
251 wait- hints databases.
253 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
254 we treated them as item separators when parsing for a list item, but they
255 need to be protected by the doublequotes. While there, add handling for
258 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
259 Found and fixed by Jasen Betts. No testcase for this as my usual text
260 editor insists on emitting only valid UTF-8.
262 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
263 a null-indirection SIGSEGV for the receive process.
265 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
266 -bh use. Internal checking would spot this and take a panic.
268 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
269 In 4.96 this would expand to empty.
271 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
272 certificate. Find and fix by Andreas Metzler.
274 JH/36 Add ARC info to DMARC hostory records.
276 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
277 or fakedefer. Previously the sender could discover that the message
278 had in fact been accepted.
280 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
281 sequences. Previously the input was not noted as being tainted; notably
282 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
283 bad coding of authenticators.
285 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
286 and ${tr...}. Found and diagnosed by Heiko Schlichting.
288 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
289 could be triggered by externally-supplied input. Found by Trend Micro.
292 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
293 be triggered by externally-controlled input. Found by Trend Micro.
296 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
297 be triggered by externally-controlled input. Found by Trend Micro.
300 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
301 Make the rewrite never match and keep the logging. Trust the
302 admin to be using verify=header-syntax (to actually reject the message).
304 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
307 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
313 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
314 after reception to before a subsequent reception. This should
315 mean slightly faster delivery, and also confirmation of reception
318 JH/02 Move from using the pcre library to pcre2. The former is no longer
319 being developed or supported (by the original developer).
321 JH/03 Constification work in the filters module required a major version
322 bump for the local-scan API. Specifically, the "headers_charset"
323 global which is visible via the API is now const and may therefore
324 not be modified by local-scan code.
326 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
327 sendfile() didi not account for the way the ClamAV driver code called it.
329 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
330 time check was being done for every character; replace that with one
333 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
334 sent was prefixed with a length byte.
336 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
337 RFC 5321. Previously Dovecot (at least) would log errors during
340 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
341 in 4.61 without a reason logged. Binaries will be bigger, which might
342 matter on diskspace-constrained systems, but debug is easier.
344 JH/09 Fix macro-definition during "-be" expansion testing. The move to
345 write-protected store for macros had not accounted for these runtime
346 additions; fix by removing this protection for "-be" mode.
348 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
349 handing out large-numbered file descriptors, violating the usual Unix
350 assumption (and required by Posix) that the lowest possible number will be
351 allocated by the kernel when a new one is needed. In the daemon, and any
352 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
353 are not useable for FD_SET() [and hence select()] and overwrite the stack.
354 Assorted crashes happen.
356 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
357 main-section options or in a connect ACL, the value from the first ever
358 connection was never replaced for subsequent connections. Found by
361 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
362 though only once PCRE2 was introduced: the memory accounting used under
363 debug offset allocations by an int, giving a hard trap in early startup.
364 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
366 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
367 with underbars is given. The write-protection of configuration introduced
368 in 4.95 trapped when normalisation was applied to an option not needing
371 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
373 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
374 failing to close the certificates directory, every hour or any time it
377 JH/16 Debugging initiated by an ACL control now continues through into routing
378 and transport processes. Previously debugging stopped any time Exim
379 re-execs, or for processing a queued message.
381 JH/17 The "expand" debug selector now gives more detail, specifically on the
382 result of expansion operators and items.
384 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
385 bad comparison between the option value and the name of the file to
386 be included was done, and a mismatch was wrongly identified.
387 4.88 to 4.95 are affected.
389 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
391 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
392 supplied as the db file-pair basename. Previously, if a directory
393 path was given, for example via the autoreply "once" option, the DB
394 file.pag and file.dir files would be created in that directory's
397 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
398 "taint" log_selector. These were previously deprecated.
400 JH/22 Fix static address-list lookups to properly return the matched item.
401 Previously only the domain part was returned.
403 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
404 the call into OpenSSL to send a TLS Close was being repeated; this
405 resulted in the library waiting for the peer's Close. If that was never
406 sent we waited forever. Fix by tracking send calls.
408 JH/24 The ${run} expansion item now expands its command string elements after
409 splitting. Previously it was before; the new ordering makes handling
410 zero-length arguments simpler. The old ordering can be obtained by
411 appending a new option "preexpand", after a comma, to the "run".
413 JH/25 Taint-check exec arguments for transport-initiated external processes.
414 Previously, tainted values could be used. This affects "pipe", "lmtp" and
415 "queryprogram" transport, transport-filter, and ETRN commands.
416 The ${run} expansion is also affected: in "preexpand" mode no part of
417 the command line may be tainted, in default mode the executable name
420 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
421 the the facility was not passed across execs, and only the first message
422 passed over a connection could use BDAT; any further ones using DATA.
424 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
425 uses $sending_ip_address and an interface is specified.
426 Previously any use of the local address in the EHLO name disabled
427 PIPECONNECT, the common case being to use the rDNS of it.
429 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
430 resumption. Previously verify failed because no certificate status is
431 passed on the wire for the restarted session. Fix by using the recorded
432 ocsp status of the stored session for the new connection.
434 JH/29 TLS resumption: the key for session lookup in the client now includes
435 more info that a server could potentially use in configuring a TLS
436 session, avoiding oferring mismatching sessions to such a server.
437 Previously only the server IP was used.
439 JH/30 Fix string_copyn() for limit greater than actual string length.
440 Previously the copied amount was the limit, which could result in a
441 overlapping memcpy for newly allocated destination soon after a
442 source string shorter than the limit. Found/investigated by KM.
444 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
445 close; it may be needed for a subsequent connection. This caused a
446 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
448 JH/32 Fix CHUNKING for a second message on a connection when the first was
449 rejected. Previously we did not reset the chunking-offered state, and
450 erroneously rejected the BDAT command. Investigation help from
453 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
454 an empty address. Previously the expansion returned an error.
456 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
457 proxy. Previously these were misparsed, leading to paniclog entries.
463 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
464 related applications. Previously an "H" was used where available info
465 says that "M" should be, so change to match.
467 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
468 as arguments, so an implementation trying to copy these into a local
469 buffer was taking a taint-enforcement trap. Fix by using dynamically
470 created buffers. Similar fix for radius expansion condition.
472 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
473 reasonable, eg. to count headers. Fix by using dynamically created
474 buffers rather than a local. Do similar fixes for ACL actions "dcc",
475 "log_reject_target", "malware" and "spam"; the arguments are expanded
476 so could be handling tainted values.
478 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
479 broken the (no-op) support for this sendmail command. Restore it
480 to doing nothing, silently, and returning good status.
482 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
483 record path was given (or the default used) without a leading directory
484 path, an error occurred on trying to open it. Use the transport's working
487 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
488 transport. Previously it was the name on the DNS A-record; use instead
489 the head of the CNAME chain leading there (if there is one). This seems
490 to align better with RFC 6125.
492 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
493 smtp_accept_max_per_host allocated resources which were not released
494 when the limit was exceeded. This eventually crashed the daemon. Fix
495 by adding a release action in that path.
497 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
498 expanded; previously using tainted values was rejected. Fix by using
499 dynamically-created buffers.
501 JH/09 Relax restrictions on ACL verify condition needing access to message
502 headers. Previously they were only permitted in data and non-smtp ACLs;
503 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
504 not_blind, header_sender and header_names_ascii verification.
506 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
507 Previously a macro used one argument twice; when called with the
508 argument as an expression having side-effects, incorrect operation
509 resulted. Use an inlineable function.
511 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
512 held open for a verify callout. Previously this wan not accounted for
513 and a corrupt onward SMTP conversation resulted.
515 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
516 Previously they were embedded naively in the constructed address; when
517 needed, strip the quoting and quote the entire local_part.
518 Also make the inbound_srs expansion condition handle quoting.
520 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
521 excluded, not matching the documentation.
523 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
524 was given for the sqlite_dbfile a trap resulted.
526 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
527 "name" argument resulted in a trap. There is no reason to disallow such;
528 this was a coding error.
530 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
531 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
532 spent suspended, ignoring the POSIX definition. Previously we assumed
533 it did and a constant offset from real time could be used as a correction.
534 Change to using the same clock source for the start-of-message and the
535 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
536 exists, just to get a clock slightly more aligned to reality.
538 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
539 RFC says it is optional some validators care. The missing char was not
540 intended but triggered by a line-wrap alignment. Discovery and fix by
541 Guillaume Outters, hacked on by JH.
543 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
544 name being quoted was tainted a trap would be taken. Fix by using
545 dynamically created buffers. The routine could have been called by a
546 rewrite with the "h" flag, by using the "-F" command-line option, or
547 by using a "name=" option on a control=submission ACL modifier.
549 JH/19 SPF: change the Authentication-Results expansion component to give
550 smtp.helo when the sender domain is empty. Previously it gave
553 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
554 not in 127.0.0.0/8 to help in spotting list domains taken over by a
555 domain-parking registrar.
557 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
558 Previously when a whitespace character was specified it was not inserted
559 after removing the newline.
561 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
562 the domain part of the recipient address. This overrides any tls_sni
563 option set, which was previously used.
565 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
568 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
569 is_tainted() had an off-by-one error in the overenthusiastic direction.
570 Find and fix by Gavan. Although NetBSD is not a supported platform for
571 4.94 this bug could affect other platforms.
573 PP/01 Fix default prime selection to be consistent.
574 One path used ike23 still, instead of exim.dev.20160529.3; now both
575 execution flows will use the same DH primes (currently
576 exim.dev.20160529.3).
578 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
579 option in smtp transport, to match the documentation. Previously
580 verification was not being done in some cases where it should have been.
582 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
583 than one server was defined and depending on the platform memory layout
584 details, an internal consistency trap could be hit while walking the list
587 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
588 files. The value is used by the authresults expansion item. Previously
589 if this was used in a router or transport, a crash could result.
591 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
592 have a sufficient incidence of EINTR returns from select that an
593 interaction with other operations done by the main daemon loop exposed
594 a bug in the error-handling. This was benign apart from the log
597 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
598 for consistency with delivered "=>" and failed "**" lines. While we're
599 there, handle PRX and TFO.
601 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
602 applied. This resulted, if any header-line rewrite rules were configured,
603 in a panic-log triggerable by sending a message with a long address in
604 a header. Fix by increasing the arbitrary limit to larger than a single
605 (dewrapped) 5322 header line maximum size.
607 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
608 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
609 RFC 6648 which deprecates X- options in protocols as a general practice.
610 Changeover between the implementations is handled by the mechanisms
613 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
614 a remote which uses callout/hold. Previously the local was queued.
616 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
619 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
620 log line. The asterisk is now consistently appended to the remote IP
621 (and port, if given), and will also be provided on defer and fail log
622 lines. Previously it could be placed on the local IP if that was being
623 logged, and was only provided on delivery lines.
625 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
627 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
628 authenticator client_send option. Previously the next char, after a pair
629 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
630 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
631 documentation. There is still no way to get a leading ^ immediately
632 after a NUL (ie. for the password of a PLAIN method authenticator.
634 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
635 files. For bad sizes read, delete the record and whine to paniclog.
637 JH/38 When logging an AUTH failure, as server, do not include sensitive
638 information. Previously, the credentials would be included if given
639 as part of the AUTH command line and an ACL denied authentication.
641 JH/39 Bug 2691: fix $local_part_data. When the matching list element
642 referred to a file, bad data was returned. This likely also affected
645 JH/40 The gsasl authenticator now supports caching of the salted password
646 generated by the client-side implementation. This required the addition
647 of a new variable: $auth4.
649 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
650 left undeleted; the attempt to re-create it then failed - resulting in
651 the usual "SIGHUP tp have daemon reload configuration" to not work.
652 This affected any platform not supporting "abstract" Unix-domain
653 sockets (i.e. not Linux).
655 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
656 recipients" response to RCPT in a later response, with a 250. The
657 previous coding assumed this would not happen, and under PIPELINING
658 would result in both lost and duplicate recipients for a message.
660 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
661 Previously the weighting was incorrectly applied. Similar fix for socks
662 proxies. Found and fixed by Heiko Schlichting.
664 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
665 not handle sub-lists included using the +namedlist syntax. While
666 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
667 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
670 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
671 Previously the permanent pool was used, so the sore could not be freed.
672 This meant a connection with many messages would use continually-growing
675 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
676 per-pool so as not to waste too much space. Previously a constant size
677 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
678 DOS attacks harder. The cost is wasted memory use in the larger blocks.
680 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
681 same space repeatedly, and a smaller process footprint.
683 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
684 for GnuTLS. Previously the name started "%s" which, while not a bug,
685 looked as if if might be one.
687 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
688 "continued connection") make the $proxy_* variables available. Previously
689 the information was not passed across the exec() call for subsequent
690 transport executions. This also mean that the log lines for the
691 messages can show the proxy information.
693 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
694 receive time. With modern systems the difference is significant.
695 The historical behaviour can be restored by disabling (a new) log_selector
696 "queue_time_exclusive".
698 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
699 content) were specifically tested for. Now, also cover expansions
700 resulting in ACL names and inline ACL content.
702 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
703 making it unusable in complex expressions.
705 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
706 fail with a taint-check on the spoolfile name, and leave the message
709 HS/01 Enforce absolute PID file path name.
711 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
713 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
715 PP/02 Bug 2643: Correct TLS DH constants.
716 A missing NUL termination in our code-generation tool had led to some
717 incorrect Diffie-Hellman constants in the Exim source.
718 Reported by kylon94, code-gen tool fix by Simon Arlott.
720 PP/03 Impose security length checks on various command-line options.
721 Fixes CVE-2020-SPRSS reported by Qualys.
723 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
724 better. Reported by Qualys.
726 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
727 providing a particularly obnoxious sender full name.
730 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
732 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
735 PP/08 Change default for recipients_max from unlimited to 50,000.
737 PP/09 Fix security issue with too many recipients on a message (to remove a
738 known security problem if someone does set recipients_max to unlimited,
739 or if local additions add to the recipient list).
740 Fixes CVE-2020-RCPTL reported by Qualys.
742 PP/10 Fix security issue in SMTP verb option parsing
743 Fixes CVE-2020-EXOPT reported by Qualys.
745 PP/11 Fix security issue in BDAT state confusion.
746 Ensure we reset known-good where we know we need to not be reading BDAT
747 data, as a general case fix, and move the places where we switch to BDAT
748 mode until after various protocol state checks.
749 Fixes CVE-2020-BDATA reported by Qualys.
751 HS/03 Die on "/../" in msglog file names
753 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
754 the Exim runtime user are allowed to create files.
756 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
759 QS/03 When reading the output from interpreted forward files we do not
760 pass the pipe between the parent and the interpreting process to
761 executed child processes (if any).
763 QS/04 Always die if requested from internal logging, even is logging is
766 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
767 the API; compilation noo longer completes with DMARC support included.
768 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
769 been reported on other platforms.
771 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
773 JH/56 Make the majority of info read from config files readonly, for defence-in-
774 depth against exploits. Suggestion by Qualys.
775 Not supported on Solaris 10.
777 JH/57 Fix control=fakreject for a custom message containing tainted data.
778 Previously this resulted in a log complaint, due to a re-expansion present
779 since fakereject was originally introduced.
781 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
782 modifier was given, a loop resulted.
784 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
785 pipelined SMTP command followed the BDAT LAST then it would be
786 incorrectly treated as part of the message body, causing a verification
789 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
790 require looser validation rules than those for 821-level addresses,
791 which only permit IP addresses.
797 JH/01 Avoid costly startup code when not strictly needed. This reduces time
798 for some exim process initialisations. It does mean that the logging
799 of TLS configuration problems is only done for the daemon startup.
801 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
803 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
804 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
807 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
808 DKIM signing is being done. Previously a transport_filter would always
809 disable CHUNKING, falling back to traditional DATA.
811 JH/05 Regard command-line recipients as tainted.
813 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
815 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
816 PAM library frees one of the arguments given to it, despite the
817 documentation. Therefore a plain malloc must be used.
819 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
820 on-stack buffers were used, resulting in a taint trap when DSN information
821 copied from a received message was written into the buffer.
823 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
824 the ordering of its ARC headers. This caused a crash.
826 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
827 a new record was being constructed with information from the peer, a trap
830 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
831 installation would get error messages from DMARC verify, when it hit the
832 nonexistent file indicated by the default. Distros wanting DMARC enabled
833 should both provide the file and set the option.
834 Also enforce no DMARC verification for command-line sourced messages.
836 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
837 could, depending on the platform, hang at the STARTTLS response.
839 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
840 message on a connection. Previously if one message had ARC headers and
841 the following one did not, a crash could result when adding an
842 Authentication-Results: header.
844 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
845 Exim main code and Exim-related utities. The introduction of taint
846 tracking also did many adjustments to string handling. Since then, eximon
847 frequently terminated with an assert failure.
849 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
850 check for 452 responses. This slightly helps the inefficieny of doing
851 a large alias-expansion into a recipient-limited target. The max_rcpt
852 transport option still applies (and at the current default, will override
853 the new feature). The check is done for either cause of synch, and forces
854 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
855 connection. The new facility is not tunable at this time.
857 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
858 library live data was being used, so the results became garbage. Make
859 copies while it is still usable.
861 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
862 on delivery deferred (==) and failed (**) lines (if a delivery was
863 attemtped). Previously it was only on completion (=>) lines.
865 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
866 for the expansion of the server_scram_iter and server_scram_salt options.
868 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
869 are now specifically given a NO_DATA response without hitting the system
870 resolver. The library goes on to do the now-standard TXT lookup.
871 Use of dnsdb lookups is not affected.
873 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
874 only retrieve the errormessage once. Previously two calls to dlerror()
875 were used, and the second one (for mainlog/paniclog) retrieved null
878 JH/20 Taint checking: disallow use of tainted data for
879 - the appendfile transport file and directory options
880 - the pipe transport command
881 - the autoreply transport file, log and once options
882 - file names used by the redirect router (including filter files)
884 - paths used by single-key lookups
885 Previously this was permitted.
887 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
888 adjusted the size of a major service buffer; this failed because the
889 buffer was in use at the time. Change to a compile-time increase in the
890 buffer size, when this authenticator is compiled into exim.
892 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
893 previous fast-mode was untenable in the face of glibs using mmap to
894 support larger malloc requests.
896 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
897 New values supported, if defined on system where compiled:
898 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
899 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
901 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
902 running a limited number of proceses in parallel, a benefit is gained. The
903 amount varies with the platform hardware and load. The use of the option
904 queue_run_in_order means we cannot do this, as ordering becomes
907 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
908 had introduced a string-copy (for ensuring NUL-termination) which was not
909 appropriate for that case, which can include embedded NUL bytes in the
910 block of data. Investigation showed the copy to actually be needless, the
911 data being length-specified.
913 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
914 done during a receiving connection, and both used TLS, global info was
915 used rather than per-connection info for tracking the state of data
916 queued for transmission. This could result in a connection hang.
918 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
919 Previously, when delivering serveral messages down a single connection
920 only the first would provide a SIZE. This was due to the size information
921 not being properly tracked.
923 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
924 TAI (at 37 seconds currently), pretend to be in UTC for time-related
925 expansion and logging. Previously, spurious values such as a future
926 minute could be seen.
928 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
929 it could crash from a null-deref. This could also affect the
930 ${addresses: } operator and ${readsock } item.
932 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
933 message following a mime one, the variable was not reset.
935 JH/30 When an pipelined-connect fails at the first response, assume incorrect
936 cached capability (perhaps the peer reneged?) and immediately retry in
939 JH/31 Fix spurious detection of timeout while writing to transport filter.
941 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
942 an attempt to copy the string was made before checking it.
944 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
945 taint of the lookup key was maintained; we now regard the presence in the
946 filesystem as sufficient validation.
948 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
949 argument is supplied.
951 JH/35 The dsearch lookup now requires that the directory is an absolute path.
952 Previously this was not checked, and nonempty relative paths made an
953 access under Exim's current working directory.
955 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
956 Previously no event was raised.
958 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
959 parameter supplied by the sender MAIL FROM command. Previously it was
960 ignored, and only the check_spool_space option value for the required
963 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
964 the size of the signing public-key. Previously it was instead giving
965 the size of the signature hash.
967 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
968 the default. See the (new) dkim_verify_min_keysizes option.
970 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
971 and an ACL use a lookup for checking either the local_part or domain,
972 stale data could be accessed. Ensure that variable references are
973 dropped between messages.
975 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
976 by the client was not checked as pointing within response data before
977 being used. A malicious client could thus cause an out-of-bounds read and
978 possibly gain authentication. Fix by adding the check.
980 JH/42 Internationalisation: change the default for downconversion in the smtp
981 transport to be "if needed". Previously it was "as previously set" for
982 the message, which usually meant "if needed" for message-submission but
983 "no" for everything else. However, MTAs have been seen using SMTPUTF8
984 even when the envelope addresses did not need it, resulting in forwarding
985 failures to non-supporting MTAs. A downconvert in such cases will be
986 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
987 transport. The change does mean that addresses needing conversion will
988 be converted when previously a delivery failure would occur.
990 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
991 response was received it would be used unchecked in a fail-DSN, violating
992 standards on line-length limits. Truncate if needed.
994 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
995 doesn't work. (Additionally add a new main config option to configure the
1002 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1003 side, to decode a TLS 1.3 packet capture.
1005 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1006 Previously the default library behaviour applied, sending two, each in
1007 its own TCP segment.
1009 JH/03 Debug output for ACL now gives the config file name and line number for
1012 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1014 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1016 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1017 buffer overrun for (non-chunking) other transports.
1019 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1020 TLS1.3, means that a server rejecting a client certificate is not visible
1021 to the client until the first read of encrypted data (typically the
1022 response to EHLO). Add detection for that case and treat it as a failed
1023 TLS connection attempt, so that the normal retry-in-clear can work (if
1024 suitably configured).
1026 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1027 and/or domain. Found and fixed by Jason Betts.
1029 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1030 configuration). If a CNAME target was not a wellformed name pattern, a
1033 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1034 the OS reports them interleaved with other addresses.
1036 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1037 used both for input and for a verify callout, both encrypted, SMTP
1038 responses being sent by the server could be lost. This resulted in
1039 dropped connections and sometimes bounces generated by a peer sending
1042 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1043 string. Previously it was possible to cause undefined behaviour in a
1044 library routine (usually a crash). Found by "zerons".
1046 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1049 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1050 API was removed, so update to use the newer ones.
1052 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1053 any timeout set, is taking a long time. Previously we would hang on to a
1054 rotated logfile "forever" if the input was arriving with long gaps
1055 (a previous attempt to fix addressed lack, for a long time, of initial
1058 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1059 shared (NFS) environment. The length of the tempfile name is now
1060 4 + 16 ("hdr.$message_exim_id") which might break on file
1061 systems which restrict the file name length to lower values.
1062 (It was "hdr.$pid".)
1064 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1065 shared (NFS) environment.
1067 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1068 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1071 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1072 on some platforms for bit 31.
1074 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1075 to changes apparently associated with TLS1.3 handling some of the APIs
1076 previously used were either nonfunctional or inappropriate. Strings
1077 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1078 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1079 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1080 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1081 and the use of specific cipher names in the encrypted= ACL condition.
1083 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1085 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1086 verification result was not updated unless hosts_require_ocsp applied.
1088 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1089 queue_list_requires_admin set to false, non-admin users were denied the
1092 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1093 directory-of-certs mode. Previously they were advertised despite the
1096 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1097 A single TCP connection by a client will now hold a TLS connection open
1098 for multiple message deliveries, by default. Previously the default was to
1101 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1102 default. If built with the facility, DANE will be used. The facility
1103 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1105 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1106 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1107 must be defined and you must still, unless you define DISABLE_TLS, manage
1108 the the include-dir and library-file requirements that go with that
1109 choice. Non-TLS builds are still supported.
1111 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1114 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1115 default. If the platform supports and has the facility enabled, it will
1116 be requested on all coneections.
1118 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1119 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1121 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1123 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1124 success-DSN messages. Previously the From: header was always the default
1125 one for these; the option was ignored.
1127 JH/28 Fix the timeout on smtp response to apply to the whole response.
1128 Previously it was reset for every read, so a teergrubing peer sending
1129 single bytes within the time limit could extend the connection for a
1130 long time. Credit to Qualsys Security Advisory Team for the discovery.
1132 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1133 delivery address, which leaked information of the results of local
1134 forwarding. Change to the original envelope recipient address, per
1137 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1138 requested. Previously not bounce was generated and a log entry of
1139 error ignored was made.
1141 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1143 JH/32 Introduce a general tainting mechanism for values read from the input
1144 channel, and values derived from them. Refuse to expand any tainted
1145 values, to catch one form of exploit.
1147 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1148 was unused and the unexpanded text used for the test. Found and
1149 fixed by Ruben Jenster.
1151 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1152 an attempt to use a TLS library read routine dereffed a nul pointer,
1155 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1156 them in our smtp response.
1158 JH/36 Have the general router option retry_use_local_part default to true when
1159 any of the restrictive preconditions are set (to anything). Previously it
1160 was only for check_local user. The change removes one item of manual
1161 configuration which is required for proper retries when a remote router
1162 handles a subset of addresses for a domain.
1164 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1165 link count into consideration.
1167 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1168 caused the extension of big_buffer, the following lines were ignored.
1170 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1171 accordance with RFC 2308. Previously there was no expiry, so a longlived
1172 receive process (eg. due to ACL delays) versus a short SOA value could
1175 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1177 JH/39 Promote DMARC support to mainline.
1179 JH/40 Bug 2452: Add a References: header to DSNs.
1181 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1182 parameters. The relevant library call is documented as "Deprecated: This
1183 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1184 3.6.0, DH parameters are negotiated following RFC7919."
1186 HS/06 Change the default of dnssec_request_domains to "*"
1188 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1189 carried on and emitted a BDAT command, even when PIPELINING was not
1192 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1193 buffer was used for the filename, resulting in a trap when tainted
1194 arguments (eg. $domain) were used.
1196 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1197 recommended to avoid a possible server-load attack. The feature can be
1198 re-enabled via the openssl_options main cofiguration option.
1200 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1201 for version 4.90 - adding a "more data" boolean to the arguments.
1202 Bumped the ABI version number also, this having been missed previously;
1203 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1204 respect of usage of smtp_printf() by either local_scan code or libraries
1205 accessed via the ${dlfunc } expansion item. Both will need coding
1206 adjustment for any calls to smtp_printf() to match the new function
1207 signature; a FALSE value for the new argument is always safe.
1209 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1210 the file-offset (which the Linux syscall does, and exim expects); this
1211 resulted in an indefinite loop.
1213 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1214 to do ARC verification. The Authentication-Results: header line added
1215 by the configuration then had no ARC item.
1217 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1218 an incorrect mode was used for reading the result, resulting in it being
1225 JH/01 Remove code calling the customisable local_scan function, unless a new
1226 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1228 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1229 non-signal-safe functions being used.
1231 JH/03 Bug 2269: When presented with a received message having a stupidly large
1232 number of DKIM-Signature headers, disable DKIM verification to avoid
1233 a resource-consumption attack. The limit is set at twenty.
1235 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1236 report of oldest_pass in ${authres } in consequence, and separate out
1237 some descriptions of reasons for verification fail.
1239 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1240 files in the spool were present and unlocked. A queue-runner could spot
1241 them, resulting in a duplicate delivery. Fix that by doing the unlock
1242 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1243 add more error-checking on spoolfile handling while that code is being
1246 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1247 No known attacks, no CVE, this is defensive hardening.
1249 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1250 a queue-runner could start a delivery while other operations were ongoing.
1251 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1252 Found and investigated by Tim Stewart. Fix by using the open message data
1253 file handle rather than opening another, and not locally closing it (which
1254 releases a lock) for that case, while creating the temporary .eml format
1255 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1257 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1258 $sender_verify_failure/$recipient_verify_failure to "random".
1260 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1263 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1264 Previously this would segfault.
1266 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1269 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1270 like zero, since the resolver should be doing this for us, But we need one
1271 as a CNAME but no MX presence gets the CNAME returned; we need to check
1272 that doesn't point to an MX to declare it "no MX returned" rather than
1273 "error, loop". A new main option is added so the older capability of
1274 following some limited number of chain links is maintained.
1276 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1278 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1279 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1280 now-deprecated earlier definitions used only specified the range up to TLS
1281 1.2 (in the older-version library docs).
1283 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1285 JH/15 Rework TLS client-side context management. Stop using a global, and
1286 explicitly pass a context around. This enables future use of TLS for
1287 connections to service-daemons (eg. malware scanning) while a client smtp
1288 connection is using TLS; with cutthrough connections this is quite likely.
1290 JH/16 Fix ARC verification to do AS checks in reverse order.
1292 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1294 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1295 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1296 Previously the "utf8" would be re-prepended for every additional message.
1298 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1299 Previously thery were accepted, resulting in issues when attempting to
1300 forward messages to a non-supporting MTA.
1302 PP/02 Let -n work with printing macros too, not just options.
1304 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1305 one parent address was copied, and bogus data was used at delivery-logging
1306 time. Either a crash (after delivery) or bogus log data could result.
1307 Discovery and analysis by Tim Stewart.
1309 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1310 Previously if the string ended mid-character, we did not insert the
1311 promised '?' replacement.
1313 PP/04 Documentation: current string operators work on bytes, not codepoints.
1315 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1316 should pack well giving a smaller memory footprint so better caching and
1317 therefore performance. Group the declarations where this can't be done so
1318 that the byte-sized flag variables are not interspersed among pointer
1319 variables, giving a better chance of good packing by the compiler.
1321 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1322 non-null, to avoid issues with sites running BATV. Previously reports were
1323 sent with an empty envelope sender so looked like bounces.
1325 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1326 The ignore_error flag wasn't being returned from the filter subprocess so
1327 was not set for later routers. Investigation and fix by Matthias Kurz.
1329 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1330 and a msg:complete for the whole, when a message is manually removed using
1331 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1333 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1334 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1335 hope that is portable enough.
1337 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1338 requiring a known-CA anchor certificate; make it now rely entirely on the
1339 TLSA as an anchor. Checking the name on the leaf cert against the name
1340 on the A-record for the host is still done for TA (but not for EE mode).
1342 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1343 included in delivery lines for non-proxied connections, when compiled with
1344 SUPPORT_SOCKS and running with proxy logging enabled.
1346 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1347 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1348 move the existing event to fire before the normal logging of message
1349 failure so that custom logging is bracketed by normal logging.
1351 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1352 msg:fail:internal event. Developement by Matthias Kurz.
1354 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1355 far too small for todays use of crypto signatures stored there. Go all
1356 the way to the max DNS message size of 64kB, even though this might be
1357 overmuch for IOT constrained device use.
1359 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1360 copy a string over itself. The library routine is documented as not
1361 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1363 JH/32 For main options check_spool_space and check_inode_space, where the
1364 platform supports 64b integers, support more than the previous 2^31 kB
1365 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1366 the previous G, M, k.
1368 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1369 $authenticated_fail_id variable on authentication failure. Previously
1372 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1373 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1374 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1375 GNUTLS_SEC_PARAM_MEDIUM.
1377 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1378 side. Previously we would continue as if no SNI had been received.
1380 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1381 "<" character, which should be a list-separator specification, we walked
1382 off past the nul-terimation.
1384 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1385 causes) even when the retry time is not yet met. Previously they were
1386 not, meaning that when (say) an account was over-quota and temp-rejecting,
1387 and multiple senders' messages were queued, only one sender would get
1388 notified on each configured delay_warning cycle.
1390 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1392 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1393 cipher-suites, an error can be left on the stack even for a succeeding
1394 accept; this results in impossible error messages when a later operation
1397 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1398 return error codes indicating retry. Under TLS1.3 this becomes required.
1400 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1401 it only wrote the new authenticators, resulting in a lack of tracking of
1402 peer changes of ESMTP extensions until the next cache flush.
1404 JH/41 Fix the loop reading a message header line to check for integer overflow,
1405 and more-often against header_maxsize. Previously a crafted message could
1406 induce a crash of the recive process; now the message is cleanly rejected.
1408 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1409 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1415 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1416 When redis_servers is set to a list of > 1 element, and the Redis servers
1417 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1418 case of MOVED into a DEFER case instead, thus moving the query onto the
1419 next server in the list. For a cluster of N elements, all N servers must
1420 be defined in redis_servers.
1422 GF/02 Catch and remove uninitialized value warning in exiqsumm
1423 Check for existence of @ARGV before looking at $ARGV[0]
1425 JH/01 Replace the store_release() internal interface with store_newblock(),
1426 which internalises the check required to safely use the old one, plus
1427 the allocate and data copy operations duplicated in both (!) of the
1428 extant use locations.
1430 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1431 modifier. This matches the restriction on the commandline.
1433 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1434 Previously only the last row was returned.
1436 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1437 we assumed that tags in the header were well-formed, and parsed the
1438 element content after inspecting only the first char of the tag.
1439 Assumptions at that stage could crash the receive process on malformed
1442 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1443 While running the DKIM ACL we operate on the Permanent memory pool so that
1444 variables created with "set" persist to the DATA ACL. Also (at any time)
1445 DNS lookups that fail create cache records using the Permanent pool. But
1446 expansions release any allocations made on the current pool - so a dnsdb
1447 lookup expansion done in the DKIM ACL releases the memory used for the
1448 DNS negative-cache, and bad things result. Solution is to switch to the
1449 Main pool for expansions.
1450 While we're in that code, add checks on the DNS cache during store_reset,
1451 active in the testsuite.
1452 Problem spotted, and debugging aided, by Wolfgang Breyha.
1454 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1455 When none of the hosts presented to a transport match an already-open
1456 connection, close it and proceed with the list. Previously we would
1457 queue the message. Spotted by Lena with Yahoo, probably involving
1460 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1461 Previously a spurious "250 OK id=" response was appended to the proper
1464 JH/08 The "support for" informational output now, which built with Content
1465 Scanning support, has a line for the malware scanner interfaces compiled
1466 in. Interface can be individually included or not at build time.
1468 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1469 by the template makefile "src/EDITME". The "STREAM" support for an older
1470 ClamAV interface method is removed.
1472 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1473 rows affected is given instead).
1475 JH/11 The runtime Berkeley DB library version is now additionally output by
1476 "exim -d -bV". Previously only the compile-time version was shown.
1478 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1479 SMTP connection. Previously, when one had more recipients than the
1480 first, an abortive onward connection was made. Move to full support for
1481 multiple onward connections in sequence, handling cutthrough connection
1482 for all multi-message initiating connections.
1484 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1485 routers. Previously, a multi-recipient message would fail to match the
1486 onward-connection opened for the first recipient, and cause its closure.
1488 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1489 a timeout on read on a GnuTLS initiating connection, resulting in the
1490 initiating connection being dropped. This mattered most when the callout
1491 was marked defer_ok. Fix to keep the two timeout-detection methods
1494 JH/15 Relax results from ACL control request to enable cutthrough, in
1495 unsupported situations, from error to silently (except under debug)
1496 ignoring. This covers use with PRDR, frozen messages, queue-only and
1499 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1501 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1502 metadata, resulting in a crash in free().
1504 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1505 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1506 Broken also in d185889f4, with init system revamp.
1508 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1509 Previously we abruptly closed the connection after reading a malware-
1510 found indication; now we go on to read the "scan ok" response line,
1513 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1514 ACL. Previously, a crash would result.
1516 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1517 macro text after a replacement (previously it was only once per line) and
1518 by skipping builtin macros when searching for an uppercase lead character.
1520 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1521 for the build is renamed.
1523 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1524 was allocated for every new TLS startup, meaning one per message. Fix
1525 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1527 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1528 reported the original. Fix to report (as far as possible) the ACL
1529 result replacing the original.
1531 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1532 OpenSSL. Certificate information is loaded for every new TLS startup,
1533 and the resources needed to be freed.
1535 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1537 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1540 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1541 DATA response info to the (existing) per-recipient response info for
1542 the "C=" log element. It can have useful tracking info from the
1543 destination system. Patch from Simon Arlott.
1545 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1546 length value. Previously this would segfault.
1548 HS/02 Support Avast multiline protoocol, this allows passing flags to
1549 newer versions of the scanner.
1551 JH/28 Ensure that variables possibly set during message acceptance are marked
1552 dead before release of memory in the daemon loop. This stops complaints
1553 about them when the debug_store option is enabled. Discovered specifically
1554 for sender_rate_period, but applies to a whole set of variables.
1555 Do the same for the queue-runner and queue-list loops, for variables set
1556 from spool message files. Do the same for the SMTP per-message loop, for
1557 certain variables indirectly set in ACL operations.
1559 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1560 as a multi-recipient message from a mailinglist manager). The coding had
1561 an arbitrary cutoff number of characters while checking for more input;
1562 enforced by writing a NUL into the buffer. This corrupted long / fast
1563 input. The problem was exposed more widely when more pipelineing of SMTP
1564 responses was introduced, and one Exim system was feeding another.
1565 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1566 receiving system, and refused recipients seen by the sending system
1567 (propating to people being dropped from mailing lists).
1568 Discovered and pinpointed by David Carter.
1570 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1571 replaced by the ${authresults } expansion.
1573 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1575 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1576 allows proper process termination in container environments.
1578 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1579 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1581 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1582 and "err_perm", deprecated since 4.83 when the RFC-defined words
1583 "temperror" and "permerror" were introduced.
1585 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1586 transport-filters or DKIM-signing. The restriction was lost in the
1587 consolidation of verify-callout and delivery SMTP handling.
1588 Extend the restriction to also cover ARC-signing.
1590 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1591 in defer=pass mode supply a 450 to the initiator. Previously the message
1594 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1595 tls_require_ciphers is used as before.
1597 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1598 "pass_unscanned". Only tmpfails from the scanner are written to
1599 the paniclog, as they may require admin intervention (permission
1600 denied, license issues). Other scanner errors (like decompression
1601 bombs) do not cause a paniclog entry.
1603 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1604 Previously it was possible to log spurious information in receive log
1607 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1608 triggered odd behaviour from Outlook Express clients.
1610 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1613 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1614 since the IETF WG has not yet settled on that versus the original
1615 "bare" representation.
1617 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1618 Previously the millisecond value corrupted the output.
1619 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1620 corrupted the output.
1626 JH/01 Rework error string handling in TLS interface so that the caller in
1627 more cases is responsible for logging. This permits library-sourced
1628 string to be attached to addresses during delivery, and collapses
1629 pairs of long lines into single ones.
1631 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1632 during configuration. Wildcards are allowed and expanded.
1634 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1635 This permits better logging.
1637 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1638 do not close it down and have the child transport start it up again on
1639 the passed-on TCP connection. Instead, proxy the child (and any
1640 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1641 affected: the continued delivery log lines do not have any DNSSEC, TLS
1642 Certificate or OCSP information. TLS cipher information is still logged.
1644 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1645 identical IP addresses on different listening ports. Will also affect
1648 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1649 add noisy ifdef guards to special-case this sillyness.
1650 Patch from Bernd Kuhls.
1652 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1653 than 255 are no longer allowed.
1655 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1656 Disable the session-cache too, which might reduce our load. Since we
1657 currrectly use a new context for every connection, both as server and
1658 client, there is no benefit for these.
1659 GnuTLS appears to not support tickets server-side by default (we don't
1660 call gnutls_session_ticket_enable_server()) but client side is enabled
1661 by default on recent versions (3.1.3 +) unless the PFS priority string
1664 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1665 <https://reproducible-builds.org/specs/source-date-epoch/>.
1667 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1668 the check for any unsuccessful recipients did not notice the limit, and
1669 erroneously found still-pending ones.
1671 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1672 MSG_MORE. Only in-clear (not on TLS connections).
1674 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1675 transport-filter is involved we can buffer the headers in memory for
1676 creating the signature, and read the spool data file once for the
1677 signature and again for transmission.
1679 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1680 4.77 as the kernel support then wasn't solid, having issues in 64bit
1681 mode. Now, it's been long enough. Add support for FreeBSD also.
1683 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1684 case where the routing stage had gathered several addresses to send to
1685 a host before calling the transport for the first, we previously failed
1686 to close down TLS in the old transport process before passing the TCP
1687 connection to the new process. The new one sent a STARTTLS command
1688 which naturally failed, giving a failed delivery and bloating the retry
1689 database. Investigation and fix prototype from Wolfgang Breyha.
1691 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1692 false-negatives in the check that the sender had not preempted a response
1693 or prompt from Exim (running as a server), due to that code's lack of
1694 awareness of the SMTP input buffering.
1696 PP/04 Add commandline_checks_require_admin option.
1697 Exim drops privileges sanely, various checks such as -be aren't a
1698 security problem, as long as you trust local users with access to their
1699 own account. When invoked by services which pass untrusted data to
1700 Exim, this might be an issue. Set this option in main configuration
1701 AND make fixes to the calling application, such as using `--` to stop
1704 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1705 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1706 responses to those, into a single TLS record each way (this usually means
1707 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1710 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1711 affects you only if you're dancing at the edge of the param size limits.
1712 If you are, and this message makes sense to you, then: raise the
1713 configured limit or use OpenSSL 1.1. Nothing we can do for older
1716 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1717 cmdline element to get the documented default one. Previously it was
1720 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1721 are unsuable for non-message-specific reasons - eg. connection timeout,
1722 banner-time rejection.
1724 JH/16 Fix logging of delivery remote port, when specified by router, under
1727 PP/06 Repair manualroute's ability to take options in any order, even if one
1728 is the name of a transport.
1731 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1733 JH/17 Change the list-building routines interface to use the expanding-string
1734 triplet model, for better allocation and copying behaviour.
1736 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1737 Previously it was constructed the first time a possibly-matching string
1738 was met in the configuration file input during startup; now it is done
1741 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1742 compatible one, to avoid the (poorly documented) possibility of a config
1743 file in the working directory redirecting the DB files, possibly correpting
1744 some existing file. CVE-2017-10140 assigned for BDB.
1746 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1747 cache-hot. Previously, although the result was properly cached, the
1748 initial verify call returned a defer.
1750 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1751 the main verify for receipient in uncached-mode.
1753 JH/22 Retire historical build files to an "unsupported" subdir. These are
1754 defined as "ones for which we have no current evidence of testing".
1756 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1757 if present. Previously it was ignored.
1759 JH/24 Start using specified-initialisers in C structure init coding. This is
1760 a C99 feature (it's 2017, so now considered safe).
1762 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1763 if was a fixed-sized field and bitmask ops via macros; it is now more
1766 PP/07 GitHub PR 56: Apply MariaDB build fix.
1767 Patch provided by Jaroslav Škarvada.
1769 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1770 during Coverity cleanups [4.87 JH/47]
1771 Diagnosis and fix provided by Michael Fischer v. Mollard.
1773 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1774 the right size to place the terminating semicolon on its own folded
1775 line, the header hash was calculated to an incorrect value thanks to
1776 the (relaxed) space the fold became.
1778 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1779 and confused the parent.
1781 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1782 which could crash as a result. This could lead to undeliverable messages.
1784 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1787 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1788 matters on fast-turnover and PID-randomising systems, which were getting
1789 out-of-order delivery.
1791 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1792 a possibly-overlapping copy. The symptom was that "Remote host closed
1793 connection in response to HELO" was logged instead of the actual 4xx
1796 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1797 Previously only that bufferd was discarded, resulting in SYMTP command
1800 JH/32 DKIM: when a message has multiple signatures matching an identity given
1801 in dkim_verify_signers, run the dkim acl once for each. Previously only
1802 one run was done. Bug 2189.
1804 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1805 "panic the current process" to "deliberately defer". The panic log is
1806 still written with the problem list name; the mail and reject logs now
1807 get a temp-reject line for the message that was being handled, saying
1808 something like "domains check lookup or other defer". The SMTP 451
1809 message is still "Temporary local problem".
1811 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1812 A crafted sequence of BDAT commands could result in in-use memory beeing
1813 freed. CVE-2017-16943.
1815 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1816 from SMTP input. Previously it was always done; now only done for DATA
1817 and not BDAT commands. CVE-2017-16944.
1819 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1820 to the message (such as an overlong header line). Previously this was
1821 not done and we did not exit BDAT mode. Followon from the previous item
1822 though a different problem.
1828 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1829 than -2003 did; needs libidn2 in addition to libidn.
1831 JH/02 The path option on a pipe transport is now expanded before use.
1833 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1834 Patch provided by "Björn", documentation fix added too.
1836 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1837 missing a wire-to-host endian conversion.
1839 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1840 close after a BDAT command line could be taken as a following command,
1841 giving a synch failure. Fix by only checking for synch immediately
1842 before acknowledging the chunk.
1844 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1845 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1846 macro. Patches provided by Josh Soref.
1848 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1849 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1850 by IANA as a service extension. However, John Klensin suggests that we
1853 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1854 the dkim code may be unix-mode line endings rather than smtp wire-format
1855 CRLF, so prepend a CR to any bare LF.
1857 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1858 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1860 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1861 the first were themselves being wrongly included in the feed into dkim
1862 processing; with most chunk sizes in use this resulted in an incorrect
1863 body hash calculated value.
1865 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1866 DKIM signature block, for verification. Although advised against by
1867 standards it is specifically not ruled illegal.
1869 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1871 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1872 missing a body hash (the bh= tag).
1874 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1875 It seems that HAProxy sends the Proxy Protocol information in clear and
1876 only then does a TLS startup, so do the same.
1878 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1879 TCP connections (such as for Spamd) unless the daemon successfully set
1880 Fast Open mode on its listening sockets. This fixes breakage seen on
1881 too-old kernels or those not configured for Fast Open, at the cost of
1882 requiring both directions being enabled for TFO, and TFO never being used
1883 by non-daemon-related Exim processes.
1885 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1886 endings, at least on the first header line. Try to canonify any that get
1887 past that check, despite the cost.
1889 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1890 now limited to an arbitrary five deep, while parsing addresses with the
1891 strip_excess_angle_brackets option enabled.
1893 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1894 instead leave the unprompted TLS handshake in socket buffer for the
1895 TLS library to consume.
1897 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1899 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1901 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1902 region in one operation, for speed, and this leaves assigned pointers
1903 dangling. Add checks run only under the testsuite which checks all
1904 variables at a store-reset and panics on a dangling pointer; add code
1905 explicitly nulling out all the variables discovered. Fixes one known
1906 bug: a transport crash, where a dangling pointer for $sending_ip_address
1907 originally assigned in a verify callout, is re-used.
1909 PP/06 Drop '.' from @INC in various Perl scripts.
1911 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1913 PP/08 Reduce a number of compilation warnings under clang; building with
1914 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1915 should be warning-free.
1917 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1919 HS/01 Fix portability problems introduced by PP/08 for platforms where
1920 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1922 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1923 chunk. This allows us to accept broken chunked messages. We need a more
1924 general solution here.
1926 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1927 already-broken messages in the queue.
1929 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1931 JH/19 Fix reference counting bug in routing-generated-address tracking.
1937 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1938 supports it and a size is available (ie. the sending peer gave us one).
1940 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1941 years of being deprecated). The replacements are the ACLs
1942 acl_smtp_mime and acl_not_smtp_mime.
1944 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1945 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1946 or even an in-clear connection were permitted. Now, if the host lookup
1947 was dnssec and dane was requested then the host is only used if the
1948 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1949 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1950 if one fails this test.
1951 This means that a poorly-configured remote DNS will make it incommunicado;
1952 but it protects against a DNS-interception attack on it.
1954 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1955 non-noisy when a race steals the message being considered.
1957 JH/05 If main configuration option tls_certificate is unset, generate a
1958 self-signed certificate for inbound TLS connections.
1960 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1961 in rewrites and routers.
1963 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1964 and logged a warning sing 4.83; now they are a configuration file error.
1966 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1967 (lacking @domain). Apply the same qualification processing as RCPT.
1969 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1971 JH/10 Support ${sha256:} applied to a string (as well as the previous
1974 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1975 a cutthrough deliver is pending, as we always want to make a connection.
1976 This also avoids re-routing the message when later placing the cutthrough
1977 connection after a verify cache hit.
1978 Do not update it with the verify result either.
1980 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1981 when routing results in more than one destination address.
1983 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1984 signing (which inhibits the cutthrough capability). Previously only
1985 the presence of an option was tested; now an expansion evaluating as
1986 empty is permissible (obviously it should depend only on data available
1987 when the cutthrough connection is made).
1989 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1990 the relevant preceding SMTP command did not note the pipelining mode.
1992 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1993 Previously they were not counted.
1995 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1996 as one having no matching records. Previously we deferred the message
1997 that needed the lookup.
1999 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2000 distinguished as "(=".
2002 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2003 for missing MX records. Previously it only worked for missing A records.
2005 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2007 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2008 after the data-go-ahead and data-ack. Patch from Jason Betts.
2010 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2011 even for a "none" policy. Patch from Tony Meyer.
2013 JH/22 Fix continued use of a connection for further deliveries. If a port was
2014 specified by a router, it must also match for the delivery to be
2017 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2018 When one of the recipients of a message was unsuitable for the connection
2019 (has no matching addresses), we lost track of needing to mark it
2020 deferred. As a result mail would be lost.
2022 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2024 JH/25 Decoding ACL controls is now done using a binary search; the source code
2025 takes up less space and should be simpler to maintain. Merge the ACL
2026 condition decode tables also, with similar effect.
2028 JH/26 Fix problem with one_time used on a redirect router which returned the
2029 parent address unchanged. A retry would see the parent address marked as
2030 delivered, so not attempt the (identical) child. As a result mail would
2033 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2034 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2035 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2038 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2039 Main config options check_{log,spool}_{inodes,space} are now
2040 100 inodes, 10MB unless set otherwise in the configuration.
2042 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2043 Previously it only applied to the main-section connection policy
2046 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2048 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2049 by me. Added RFC7919 DH primes as an alternative.
2051 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2052 are not in the system include path.
2054 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2055 GnuTLS, when a session startup failed (eg because the client disconnected)
2056 Exim did stdio operations after fclose. This was exposed by a recent
2057 change which nulled out the file handle after the fclose.
2059 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2060 signed directly by the cert-signing cert, rather than an intermediate
2061 OCSP-signing cert. This is the model used by LetsEncrypt.
2063 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2065 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2066 an incoming connection.
2068 HS/02 Bug 1802: Do not half-close the connection after sending a request
2071 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2072 fallback to "prime256v1".
2074 JH/34 SECURITY: Use proper copy of DATA command in error message.
2075 Could leak key material. Remotely exploitable. CVE-2016-9963.
2081 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2082 and 3.4.4 - once the server is enabled to respond to an OCSP request
2083 it does even when not requested, resulting in a stapling non-aware
2084 client dropping the TLS connection.
2086 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2087 support variable-length bit vectors. No functional change.
2089 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2090 The I= interface field on outgoing lines is now after the H= remote
2091 host field, same as incoming lines. There is a separate
2092 outgoing_interface log selector which allows you to disable the
2095 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2096 If not running log_selector +smtp_connection the mainlog would be held
2097 open indefinitely after a "too many connections" event, including to a
2098 deleted file after a log rotate. Leave the per net connection logging
2099 leaving it open for efficiency as that will be quickly detected by the
2100 check on the next write.
2102 HS/01 Bug 1671: Fix post transport crash.
2103 Processing the wait-<transport> messages could crash the delivery
2104 process if the message IDs didn't exist for some reason. When
2105 using 'split_spool_directory=yes' the construction of the spool
2106 file name failed already, exposing the same netto behaviour.
2108 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2109 mime_regex ACL conditions.
2111 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2112 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2113 to HELO, local diagnostic string.
2115 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2116 log line to debug. Even when configured with a tls authenticator many
2117 client connections are expected to not authenticate in this way, so
2118 an authenticate fail is not an error.
2120 HS/02 Add the Exim version string to the process info. This way exiwhat
2121 gives some more detail about the running daemon.
2123 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2124 matter for fast-change records such as DNSBLs.
2126 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2127 retry record, even if constant. There may be multiple transports with
2128 different interface settings and the retry behaviour needs to be kept
2131 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2133 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2135 JH/10 Bug 840: fix log_defer_output option of pipe transport
2137 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2138 hosts_randomize. This matters a lot when combined with mua_wrapper.
2140 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2141 ${quote_pgsql:<string>} operator.
2143 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2145 JH/14 Change header file name for freeradius-client. Relevant if compiling
2146 with Radius support; from the Gentoo tree and checked under Fedora.
2148 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2150 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2151 option transport_filter, meaning no filtering.
2153 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2155 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2156 defaults to "*" (all hosts). The variable is now available when not built
2157 with TLS, default unset, mainly to enable keeping the testsuite sane.
2158 If a server certificate is not supplied (via tls_certificate) an error is
2159 logged, and clients will find TLS connections fail on startup. Presumably
2160 they will retry in-clear.
2161 Packagers of Exim are strongly encouraged to create a server certificate
2162 at installation time.
2164 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2165 with the $config_file variable.
2167 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2168 in transport context, after the attempt, and per-recipient. The latter type
2169 is per host attempted. The event data is the error message, and the errno
2170 information encodes the lookup type (A vs. MX) used for the (first) host,
2171 and the trailing two digits of the smtp 4xx response.
2173 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2174 to write to mainlog (or rejectlog, paniclog) in the window between file
2175 creation and permissions/ownership being changed. Particularly affects
2176 installations where exicyclog is run as root, rather than exim user;
2177 result is that the running daemon panics and dies.
2179 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2181 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2182 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2183 "pri" and "weight". Note that the previous implicit priority given by the
2184 list order is no longer honoured.
2186 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2187 for DKIM processing.
2189 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2190 by defining SUPPORT_SOCKS.
2192 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2193 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2194 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2195 variables are renamed to proxy_{local,external}_{address,port}.
2197 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2198 for a build by defining SUPPORT_I18N
2200 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2201 of the query string, and make ${quote_redis:} do that quoting.
2203 JH/29 Move Events support from Experimental to mainline, enabled by default
2204 and removable for a build by defining DISABLE_EVENT.
2206 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2208 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2209 cached by the daemon.
2211 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2212 by defining LOOKUP_REDIS. The libhiredis library is required.
2214 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2215 keys are given for lookup.
2217 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2218 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2219 only supported when built with TLS support. The PolarSSL SHA routines
2220 are still used when the TLS library is too old for convenient support.
2222 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2223 openssl_options), for security. OpenSSL forces this from version 1.1.0
2224 server-side so match that on older versions.
2226 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2227 allocation for $value could be released as the expansion processing
2228 concluded, but leaving the global pointer active for it.
2230 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2231 and to use the domains and local_parts ACL conditions.
2233 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2234 incorrectly not doubled on cutthrough transmission, hence seen as a
2235 body-termination at the receiving system - resulting in truncated mails.
2236 Commonly the sender saw a TCP-level error, and retransmitted the message
2237 via the normal store-and-forward channel. This could result in duplicates
2238 received - but deduplicating mailstores were liable to retain only the
2239 initial truncated version.
2241 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2243 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2245 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2246 we're in there, support oversigning also; bug 1309.
2248 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2250 HS/04 Add support for keep_environment and add_environment options.
2252 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2253 either intentional arithmetic overflow during PRNG, or testing config-
2256 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2257 delivery resulted in actual delivery. Cancel cutthrough before DATA
2260 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2261 rejects a recipient: pass the reject to the originator.
2263 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2264 Many were false-positives and ignorable, but it's worth fixing the
2267 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2268 for the new environment-manipulation done at startup. Move the routines
2269 from being local to tls.c to being global via the os.c file.
2271 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2272 an extract embedded as result-arg for a map, the first arg for extract
2273 is unavailable so we cannot tell if this is a numbered or keyed
2274 extraction. Accept either.
2280 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2283 JH/02 The smtp transport option "multi_domain" is now expanded.
2285 JH/03 The smtp transport now requests PRDR by default, if the server offers
2288 JH/04 Certificate name checking on server certificates, when exim is a client,
2289 is now done by default. The transport option tls_verify_cert_hostnames
2290 can be used to disable this per-host. The build option
2291 EXPERIMENTAL_CERTNAMES is withdrawn.
2293 JH/05 The value of the tls_verify_certificates smtp transport and main options
2294 default to the word "system" to access the system default CA bundle.
2295 For GnuTLS, only version 3.0.20 or later.
2297 JH/06 Verification of the server certificate for a TLS connection is now tried
2298 (but not required) by default. The verification status is now logged by
2299 default, for both outbound TLS and client-certificate supplying inbound
2302 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2305 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2306 Status Notification (bounce) messages are now MIME format per RFC 3464.
2307 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2308 under the control of the dsn_advertise_hosts option, and routers may
2309 have a dsn_lasthop option.
2311 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2312 default, modifiable by a malware= option. The list separator for
2313 the options can now be changed in the usual way. Bug 68.
2315 JH/10 The smtp_receive_timeout main option is now expanded before use.
2317 JH/11 The incoming_interface log option now also enables logging of the
2318 local interface on delivery outgoing connections.
2320 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2321 if the interface and destination host and port all match.
2323 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2326 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2327 Patch from Andrew Lewis.
2329 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2330 now supports optional time-restrictions, weighting, and priority
2331 modifiers per server. Patch originally by <rommer@active.by>.
2333 JH/16 The spamd_address main option now supports a mixed list of local
2334 and remote servers. Remote servers can be IPv6 addresses, and
2335 specify a port-range.
2337 JH/17 Bug 68: The spamd_address main option now supports an optional
2338 timeout value per server.
2340 JH/18 Bug 1581: Router and transport options headers_add/remove can
2341 now have the list separator specified.
2343 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2346 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2349 JH/21 Support for the A6 type of dns record is withdrawn.
2351 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2352 rather than the verbs used.
2354 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2355 from 255 to 1024 chars.
2357 JH/24 Verification callouts now attempt to use TLS by default.
2359 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2360 are generic router options now. The defaults didn't change.
2362 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2363 Original patch from Alexander Shikoff, worked over by JH.
2365 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2366 files if ZCAT_COMMAND is not executable.
2368 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2370 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2372 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2373 Normally benign, it bites when the pair was led to by a CNAME;
2374 modern usage is to not canonicalize the domain to a CNAME target
2375 (and we were inconsistent anyway for A-only vs AAAA+A).
2377 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2379 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2380 when evaluating $sender_host_dnssec.
2382 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2383 $sender_helo_dnssec variable.
2385 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2387 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2389 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2391 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2392 documented as working, but never had. Support all but $spam_report.
2394 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2395 added for tls authenticator.
2397 HS/03 Add perl_taintmode main config option
2403 TL/01 When running the test suite, the README says that variables such as
2404 no_msglog_check are global and can be placed anywhere in a specific
2405 test's script, however it was observed that placement needed to be near
2406 the beginning for it to behave that way. Changed the runtest perl
2407 script to read through the entire script once to detect and set these
2408 variables, reset to the beginning of the script, and then run through
2409 the script parsing/test process like normal.
2411 TL/02 The BSD's have an arc4random API. One of the functions to induce
2412 adding randomness was arc4random_stir(), but it has been removed in
2413 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2414 function when detected.
2416 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2417 cause callback expansion.
2419 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2420 syntax errors in an expansion can be treated as a string instead of
2421 logging or causing an error, due to the internal use of bool_lax
2422 instead of bool when processing it.
2424 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2425 server certificates when making smtp deliveries.
2427 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2429 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2431 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2433 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2434 Merged patch from Sebastian Wiedenroth.
2436 JH/05 Fix results-pipe from transport process. Several recipients, combined
2437 with certificate use, exposed issues where response data items split
2438 over buffer boundaries were not parsed properly. This eventually
2439 resulted in duplicates being sent. This issue only became common enough
2440 to notice due to the introduction of connection certificate information,
2441 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2443 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2444 size buffer was used, resulting in syntax errors when an expansion
2447 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2448 version 3.3.6 or later.
2450 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2451 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2452 both become "event_action", the variables become $event_name, $event_data
2453 and $event_defer_errno. There is a new variable $verify_mode, usable in
2454 routers, transports and related events. The tls:cert event is now also
2455 raised for inbound connections, if the main configuration event_action
2458 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2459 early OCSP support, but no stapling (appears to be less than 1.0.0).
2461 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2462 server certificate names available under the smtp transport option
2463 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2466 JH/10 Time-related extraction expansions from certificates now use the main
2467 option "timezone" setting for output formatting, and are consistent
2468 between OpenSSL and GnuTLS compilations. Bug 1541.
2470 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2471 encoded parameter in the incoming message. Bug 1558.
2473 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2474 include certificate info, eximon was claiming there were spoolfile
2477 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2479 JH/14 Log delivery-related information more consistently, using the sequence
2480 "H=<name> [<ip>]" wherever possible.
2482 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2483 are problematic for Debian distribution, omit them from the release
2486 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2488 JH/16 Fix string representation of time values on 64bit time_t architectures.
2491 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2492 output list separator was used.
2497 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2498 checkers that were complaining about end of non-void function with no
2501 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2502 This was a regression introduced in 4.83 by another bugfix.
2504 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2506 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2507 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2513 TF/01 Correctly close the server side of TLS when forking for delivery.
2515 When a message was received over SMTP with TLS, Exim failed to clear up
2516 the incoming connection properly after forking off the child process to
2517 deliver the message. In some situations the subsequent outgoing
2518 delivery connection happened to have the same fd number as the incoming
2519 connection previously had. Exim would try to use TLS and fail, logging
2520 a "Bad file descriptor" error.
2522 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2523 utilities have not been installed.
2525 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2526 temporary space as the ACL may create new global variables.
2528 TL/01 LDAP support uses per connection or global context settings, depending
2529 upon the detected version of the libraries at build time.
2531 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2532 to extract and use the src ip:port in logging and expansions as if it
2533 were a direct connection from the outside internet. PPv2 support was
2534 updated based on HAProxy spec change in May 2014.
2536 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2538 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2539 Properly escape header and check for NULL return.
2541 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2544 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2546 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2547 characters in header names, implemented as a verify condition.
2548 Contributed by Michael Fischer v. Mollard.
2550 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2551 results permerror and temperror. Previous values are deprecated but
2552 still accepted. In a future release, err_perm and err_temp will be
2553 completely removed, which will be a backward incompatibility if the
2554 ACL tests for either of these two old results. Patch contributed by
2555 user bes-internal on the mailing list.
2557 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2559 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2560 selectors, in both main and reject logs.
2562 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2565 JH/07 Add malware type "sock" for talking to simple daemon.
2567 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2569 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2570 routers/transports under cutthrough routing.
2572 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2573 numbers. Touch up "bool" conditional to keep the same definition.
2575 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2577 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2579 JH/12 Expand items in router/transport headers_add or headers_remove lists
2580 individually rather than the list as a whole. Bug 1452.
2582 Required for reasonable handling of multiple headers_ options when
2583 they may be empty; requires that headers_remove items with embedded
2584 colons must have them doubled (or the list-separator changed).
2586 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2587 view the policy declared in the DMARC record. Currently, $dmarc_status
2588 is a combined value of both the record presence and the result of the
2591 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2593 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2594 dnslookup router and the smtp transport (applying to the forward
2597 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2598 of ldap servers used for a specific lookup. Patch provided by Heiko
2601 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2602 New variable $lookup_dnssec_authenticated for observability.
2604 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2605 Patch submitted by Lars Timman.
2607 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2609 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2610 Requires trusted mode and valid format message id, aborts otherwise.
2611 Patch contributed by Heiko Schlichting.
2613 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2614 certextract with support for various fields. Bug 1358.
2616 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2617 is requested by default, modifiable by smtp transport option
2620 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2621 operate on certificate variables to give certificate fingerprints
2622 Also new ${sha256:cert_variable}.
2624 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2626 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2627 Christian Aistleitner.
2629 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2631 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2632 file. Patch from Wolfgang Breyha.
2634 JH/25 Expand the coverage of the delivery $host and $host_address to
2635 client authenticators run in verify callout. Bug 1476.
2637 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2638 align with daemon_smtp_ports. Bug 72.
2640 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2641 support and error reporting did not work properly.
2643 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2644 and is readable. Patch from Andrew Colin Kissa.
2646 TL/14 Enhance documentation of ${run expansion and how it parses the
2647 commandline after expansion, particularly in the case when an
2648 unquoted variable expansion results in an empty value.
2650 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2652 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2653 by renaming away from it. Addresses GH issue 15, reported by
2656 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2658 TL/15 SECURITY: prevent double expansion in math comparison functions
2659 (can expand unsanitized data). Not remotely exploitable.
2666 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2668 PP/02 Make -n do something, by making it not do something.
2669 When combined with -bP, the name of an option is not output.
2671 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2674 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2675 $sender_host_name and config options to manage this, and basic check
2678 PP/05 DSCP support for outbound connections and control modifier for inbound.
2680 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2681 (Only plugin which currently uses this is kerberos4, which nobody should
2682 be using, but we should make it available and other future plugins might
2683 conceivably use it, even though it would break NAT; stuff *should* be
2684 using channel bindings instead).
2686 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2687 name; added for Sendmail compatibility; requires admin caller.
2688 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2689 just ignore it); requires trusted caller.
2690 Also parse but ignore: -Ac -Am -X<logfile>
2693 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2695 TL/02 Add +smtp_confirmation as a default logging option.
2697 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2698 Patch by Magnus Holmgren from 2007-02-20.
2700 TL/04 Bugzilla 1281 - Spec typo.
2701 Bugzilla 1283 - Spec typo.
2702 Bugzilla 1290 - Spec grammar fixes.
2704 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2706 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2708 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2709 mailing list by by Dmitry Isaikin.
2711 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2713 JH/02 Support "G" suffix to numbers in ${if comparisons.
2715 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2717 NM/01 Bugzilla 1197 - Spec typo
2718 Bugzilla 1196 - Spec examples corrections
2720 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2722 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2723 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2726 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2727 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2729 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2730 "acl {{name}{arg}...}", and optional args on acl condition
2733 JH/05 Permit multiple router/transport headers_add/remove lines.
2735 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2737 JH/07 Avoid using a waiting database for a single-message-only transport.
2738 Performance patch from Paul Fisher. Bugzilla 1262.
2740 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2743 JH/09 Add $headers_added variable, with content from use of ACL modifier
2744 add_header (but not yet added to the message). Bugzilla 199.
2746 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2747 Pulled from Bugzilla 817 by Wolfgang Breyha.
2749 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2751 (nb: this is the same fix as in Exim 4.80.1)
2753 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2756 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2759 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2761 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2762 advertise SMTP AUTH mechanism to us, instead of a generic
2763 protocol violation error. Also, make Exim more robust to bad
2764 data from the Dovecot auth socket.
2766 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2768 When a queue runner is handling a message, Exim first routes the
2769 recipient addresses, during which it prunes them based on the retry
2770 hints database. After that it attempts to deliver the message to
2771 any remaining recipients. It then updates the hints database using
2774 So if a recipient address works intermittently, it can get repeatedly
2775 deferred at routing time. The retry hints record remains fresh so the
2776 address never reaches the final cutoff time.
2778 This is a fairly common occurrence when a user is bumping up against
2779 their storage quota. Exim had some logic in its local delivery code
2780 to deal with this. However it did not apply to per-recipient defers
2781 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2783 This change adds a proper retry rule check during routing so that the
2784 final cutoff time is checked against the message's age. We only do
2785 this check if there is an address retry record and there is not a
2786 domain retry record; this implies that previous attempts to handle
2787 the address had the retry_use_local_parts option turned on. We use
2788 this as an approximation for the destination being like a local
2789 delivery, as in LMTP.
2791 I suspect this new check makes the old local delivery cutoff check
2792 redundant, but I have not verified this so I left the code in place.
2794 TF/02 Correct gecos expansion when From: is a prefix of the username.
2796 Test 0254 submits a message to Exim with the header
2800 When I ran the test suite under the user fanf2, Exim expanded
2801 the header to contain my full name, whereas it should have added
2802 a Resent-Sender: header. It erroneously treats any prefix of the
2803 username as equal to the username.
2805 This change corrects that bug.
2807 GF/01 DCC debug and logging tidyup
2808 Error conditions log to paniclog rather than rejectlog.
2809 Debug lines prefixed by "DCC: " to remove any ambiguity.
2811 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2813 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2814 Bug spotted by Jeremy Harris; was flawed since initial commit.
2815 Would have resulted in OCSP responses post-SNI triggering an Exim
2816 NULL dereference and crash.
2818 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2820 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2821 Bug detection, analysis and fix by Samuel Thibault.
2822 Bugzilla 1331, Debian bug #698092.
2824 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2826 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2827 Server implementation by Todd Lyons, client by JH.
2828 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2829 config variable "prdr_enable" controls whether the server
2830 advertises the facility. If the client requests PRDR a new
2831 acl_data_smtp_prdr ACL is called once for each recipient, after
2832 the body content is received and before the acl_smtp_data ACL.
2833 The client is controlled by both of: a hosts_try_prdr option
2834 on the smtp transport, and the server advertisement.
2835 Default client logging of deliveries and rejections involving
2836 PRDR are flagged with the string "PRDR".
2838 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2839 fclose(). Diagnosis by Todd Lyons.
2841 PP/17 Update configure.default to handle IPv6 localhost better.
2842 Patch by Alain Williams (plus minor tweaks).
2845 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2846 This is now consistent with GnuTLS, and is now documented: the
2847 previous undocumented portable approach to treating the option as
2848 unset was to force an expansion failure. That still works, and
2849 an empty string is now equivalent.
2851 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2852 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2853 not performing validation itself.
2855 PP/20 Added force_command boolean option to pipe transport.
2856 Patch from Nick Koston, of cPanel Inc.
2858 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2861 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2863 PP/21 Fix eximon continuous updating with timestamped log-files.
2864 Broken in a format-string cleanup in 4.80, missed when I repaired the
2865 other false fix of the same issue.
2866 Report and fix from Heiko Schlichting.
2869 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2870 Report from Prashanth Katuri.
2872 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2873 It's SecureTransport, so affects any MacOS clients which use the
2874 system-integrated TLS libraries, including email clients.
2876 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2877 using a MIME ACL for non-SMTP local injection.
2878 Report and assistance in diagnosis by Warren Baker.
2880 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2882 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2884 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2885 last id that failed so it may be referenced in subsequent ACL's.
2887 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2890 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2891 ldap library initialization, allowing self-signed CA's to be
2892 used. Also properly sets require_cert option later in code by
2893 using NULL (global ldap config) instead of ldap handle (per
2894 session). Bug diagnosis and testing by alxgomz.
2896 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2897 the src/util/ subdirectory.
2899 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2900 renamed to Transport Post Delivery Action by Jeremy Harris, as
2903 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2904 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2905 redis_servers = needs to be configured which will be used by the redis
2906 lookup. Patch from Warren Baker, of The Packet Hub.
2908 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2910 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2911 hostname or reverse DNS when processing a host list. Used suggestions
2912 from multiple comments on this bug.
2914 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2916 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2917 Missed a few lines, added it to make the runtest require no keyboard
2920 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2921 contains upper case chars. Make router use caseful_local_part.
2923 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2924 support when GnuTLS has been built with p11-kit.
2930 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2932 This, or similar/improved, will also be change PP/11 of 4.82.
2938 PP/01 Handle short writes when writing local log-files.
2939 In practice, only affects FreeBSD (8 onwards).
2940 Bugzilla 1053, with thanks to Dmitry Isaikin.
2942 NM/01 Bugzilla 949 - Documentation tweak
2944 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2947 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2949 PP/02 Implemented gsasl authenticator.
2951 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2953 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2954 `pkg-config foo` for cflags/libs.
2956 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2957 with rest of GSASL and with heimdal_gssapi.
2959 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2960 `pkg-config foo` for cflags/libs for the TLS implementation.
2962 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2963 properties get this fed in as external SSF. A number of robustness
2964 and debugging improvements to the cyrus_sasl authenticator.
2966 PP/08 cyrus_sasl server now expands the server_realm option.
2968 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2969 Patch by Jeremy Harris.
2971 PP/10 Added dbmjz lookup type.
2973 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2975 PP/12 MAIL args handles TAB as well as SP, for better interop with
2976 non-compliant senders.
2977 Analysis and variant patch by Todd Lyons.
2979 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2980 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2981 Patch from Dirk Mueller <dmueller@suse.com>
2983 PP/13 tls_peerdn now print-escaped for spool files.
2984 Observed some $tls_peerdn in wild which contained \n, which resulted
2985 in spool file corruption.
2987 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2988 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2989 or write after TLS renegotiation, which otherwise led to messages
2992 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2993 as a tracking header (ie: a signed header comes before the signature).
2994 Patch from Wolfgang Breyha.
2996 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2997 comma-sep list; embedded commas doubled.
2999 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3001 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3003 Report and patch from Dmitry Banschikov.
3005 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3006 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3007 failures. We appear to now support TLS1.1+ with Exim.
3009 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3010 lets Exim select keys and certificates based upon TLS SNI from client.
3011 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3012 before an outbound SMTP session. New log_selector, +tls_sni.
3014 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3015 NULL dereference. Report and patch from Alun Jones.
3017 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3018 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3019 Not seeing resolver debug output on NetBSD, but suspect this is a
3020 resolver implementation change.
3022 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3023 Left warnings. Added "eximon gdb" invocation mode.
3025 PP/21 Defaulting "accept_8bitmime" to true, not false.
3027 PP/22 Added -bw for inetd wait mode support.
3029 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3030 locate the relevant includes and libraries. Made this the default.
3032 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3033 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3035 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3036 This may cause build issues on older platforms.
3038 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3039 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3040 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3041 Added SNI support via GnuTLS too.
3042 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3044 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3046 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3047 Applied second patch from Janne, implementing suggestion to default
3048 multiple-strings-in-record handling to match SPF spec.
3050 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3052 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3053 read-only, out of scope).
3054 Patch from Wolfgang Breyha, report from Stuart Northfield.
3056 PP/29 Fix three issues highlighted by clang analyser static analysis.
3057 Only crash-plausible issue would require the Cambridge-specific
3058 iplookup router and a misconfiguration.
3059 Report from Marcin Mirosław.
3061 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3063 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3064 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3065 As part of this, removing so much warning spew let me fix some minor
3066 real issues in debug logging.
3068 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3069 assignment on my part. Fixed.
3071 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3072 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3073 Janne Snabb (who went above and beyond: thank you).
3075 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3076 string otherwise requires a connection and a bunch more work and it's
3077 relatively easy to get wrong. Should also expose TLS library linkage
3080 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3081 64-bit ${eval} (JH/03).
3083 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3084 GNU libc to support some of the 64-bit stuff, should not lead to
3085 conflicts. Defined before os.h is pulled in, so if a given platform
3086 needs to override this, it can.
3088 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3089 protection layer was required, which is not implemented.
3090 Bugzilla 1254, patch from Wolfgang Breyha.
3092 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3093 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3094 tls_dhparam take prime identifiers. Also unbreak combination of
3095 OpenSSL+DH_params+TLSSNI.
3097 PP/39 Disable SSLv2 by default in OpenSSL support.
3103 PP/01 Solaris build fix for Oracle's LDAP libraries.
3104 Bugzilla 1109, patch from Stephen Usher.
3106 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3108 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3111 TF/02 Fix a couple more cases where we did not log the error message
3112 when unlink() failed. See also change 4.74-TF/03.
3114 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3115 lock up or crash if it happened to be inside a call to libc when it
3116 got a SIGUSR1 from exiwhat.
3118 The SIGUSR1 handler appends the current process status to the process
3119 log which is later printed by exiwhat. It used to use the general
3120 purpose logging code to do this, but several functions it calls are
3121 not safe for signals.
3123 The new output code in the SIGUSR1 handler is specific to the process
3124 log, and simple enough that it's easy to inspect for signal safety.
3125 Removing some special cases also simplifies the general logging code.
3126 Removing the spurious timestamps from the process log simplifies
3129 TF/04 Improved ratelimit ACL condition.
3131 The /noupdate option has been deprecated in favour of /readonly which
3132 has clearer semantics. The /leaky, /strict, and /readonly update modes
3133 are mutually exclusive. The update mode is no longer included in the
3134 database key; it just determines when the database is updated. (This
3135 means that when you upgrade Exim will forget old rate measurements.)
3137 Exim now checks that the per_* options are used with an update mode that
3138 makes sense for the current ACL. For example, when Exim is processing a
3139 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3140 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3141 must specify per_mail/readonly. If you omit the update mode it defaults to
3142 /leaky where that makes sense (as before) or /readonly where required.
3144 The /noupdate option is now undocumented but still supported for
3145 backwards compatibility. It is equivalent to /readonly except that in
3146 ACLs where /readonly is required you may specify /leaky/noupdate or
3147 /strict/noupdate which are treated the same as /readonly.
3149 A useful new feature is the /count= option. This is a generalization
3150 of the per_byte option, so that you can measure the throughput of other
3151 aggregate values. For example, the per_byte option is now equivalent
3152 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3154 The per_rcpt option has been generalized using the /count= mechanism
3155 (though it's more complicated than the per_byte equivalence). When it is
3156 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3157 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3158 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3159 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3160 Note that using per_rcpt with a non-readonly update mode in more than
3161 one ACL will cause the recipients to be double-counted. (The per_mail
3162 and per_byte options don't have this problem.)
3164 The handling of very low rates has changed slightly. If the computed rate
3165 is less than the event's count (usually one) then this event is the first
3166 after a long gap. In this case the rate is set to the same as this event's
3167 count, so that the first message of a spam run is counted properly.
3169 The major new feature is a mechanism for counting the rate of unique
3170 events. The new per_addr option counts the number of different
3171 recipients that someone has sent messages to in the last time period. It
3172 behaves like per_rcpt if all the recipient addresses are different, but
3173 duplicate recipient addresses do not increase the measured rate. Like
3174 the /count= option this is a general mechanism, so the per_addr option
3175 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3176 example, measure the rate that a client uses different sender addresses
3177 with the options per_mail/unique=$sender_address. There are further
3178 details in the main documentation.
3180 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3182 TF/06 Removed a few PCRE remnants.
3184 TF/07 Automatically extract Exim's version number from tags in the git
3185 repository when doing development or release builds.
3187 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3188 Bugzilla 879. Patch from Paul Fisher.
3190 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3191 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3194 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3196 PP/05 Variable $av_failed, true if the AV scanner deferred.
3197 Bugzilla 1078. Patch from John Horne.
3199 PP/06 Stop make process more reliably on build failure.
3200 Bugzilla 1087. Patch from Heiko Schlittermann.
3202 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3203 Bugzilla 1089. Patch from Heiko Schlittermann.
3205 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3206 Bugzilla 1131. Patch from Holger Weiß.
3208 PP/09 Handle IPv6 addresses with SPF.
3209 Bugzilla 860. Patch from Wolfgang Breyha.
3211 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3213 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3216 PP/11 match_* no longer expand right-hand-side by default.
3217 New compile-time build option, EXPAND_LISTMATCH_RHS.
3218 New expansion conditions, "inlist", "inlisti".
3220 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3222 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3224 PP/14 fix log_write() format string regression from TF/03.
3225 Bugzilla 1152. Patch from Dmitry Isaikin.
3231 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3233 PP/02 Harmonised TLS library version reporting; only show if debugging.
3234 Layout now matches that introduced for other libraries in 4.74 PP/03.
3236 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3238 PP/04 New "dns_use_edns0" global option.
3240 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3243 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3244 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3246 TK/01 Updated PolarSSL code to 0.14.2.
3247 Bugzilla 1097. Patch from Andreas Metzler.
3249 PP/07 Catch divide-by-zero in ${eval:...}.
3250 Fixes bugzilla 1102.
3252 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3255 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3256 format-string attack -- SECURITY: remote arbitrary code execution.
3258 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3259 time unintentionally subject to list matching rules, letting the header
3260 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3261 arbitrary string expansion). This allowed for information disclosure.
3263 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3264 INT_MIN/-1 -- value coerced to INT_MAX.
3270 NM/01 Workaround for PCRE version dependency in version reporting
3273 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3274 This fixes portability to compilers other than gcc, notably
3275 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3277 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3278 makefiles for portability to HP-UX and POSIX correctness.
3280 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3281 Also via indented variable definition in the Makefile.
3282 (Debugging by Oliver Heesakkers).
3284 PP/02 Restore caching of spamd results with expanded spamd_address.
3285 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3287 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3288 Improves build reliability. Fix from: Frank Elsner
3290 NM/02 Fix wide character breakage in the rfc2047 coding
3291 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3293 NM/03 Allow underscore in dnslist lookups
3294 Fixes bug 1026. Patch from Graeme Fowler
3296 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3297 Code patches from Adam Ciarcinski of NetBSD.
3299 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3302 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3303 is logged, to avoid truncation. Patch from John Horne.
3305 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3306 Patch from Jakob Hirsch.
3308 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3309 SQL string expansion failure details.
3310 Patch from Andrey Oktyabrski.
3312 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3313 Patch from Simon Arlott.
3315 PP/09 New lookups functionality failed to compile on old gcc which rejects
3316 extern declarations in function scope.
3317 Patch from Oliver Fleischmann
3319 PP/10 Use sig_atomic_t for flags set from signal handlers.
3320 Check getgroups() return and improve debugging.
3321 Fixed developed for diagnosis in bug 927 (which turned out to be
3324 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3325 Patch from Mark Zealey.
3327 PP/12 Bugzilla 1056: Improved spamd server selection.
3328 Patch from Mark Zealey.
3330 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3331 Based on patch from Heiko Schlittermann.
3333 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3334 Patch from Uwe Doering, sign-off by Michael Haardt.
3336 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3337 variable declaration deep within a block. Bug and patch from
3340 PP/15 lookups-Makefile IRIX compatibility coercion.
3342 PP/16 Make DISABLE_DKIM build knob functional.
3344 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3345 Patch by Simon Arlott
3347 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3348 variable argument macros. Our copy now differs from upstream.
3354 TF/01 Failure to get a lock on a hints database can have serious
3355 consequences so log it to the panic log.
3357 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3358 controlled using the smtp_confirmation log selector.
3360 TF/03 Include the error message when we fail to unlink a spool file.
3362 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3363 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3364 for maintaining out-of-tree patches for some time.
3366 PP/01 Bugzilla 139: Documentation and portability issues.
3367 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3368 Handle per-OS dynamic-module compilation flags.
3370 PP/02 Let /dev/null have normal permissions.
3371 The 4.73 fixes were a little too stringent and complained about the
3372 permissions on /dev/null. Exempt it from some checks.
3373 Reported by Andreas M. Kirchwitz.
3375 PP/03 Report version information for many libraries, including
3376 Exim version information for dynamically loaded libraries. Created
3377 version.h, now support a version extension string for distributors
3378 who patch heavily. Dynamic module ABI change.
3380 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3381 privilege escalation vulnerability whereby the Exim run-time user
3382 can cause root to append content of the attacker's choosing to
3385 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3388 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3389 If dropping privileges for untrusted macros, we disabled normal logging
3390 on the basis that it would fail; for the Exim run-time user, this is not
3391 the case, and it resulted in successful deliveries going unlogged.
3392 Fixed. Reported by Andreas Metzler.
3398 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3399 only prepend for the Resent-* case. Fixes regression introduced in
3400 Exim 4.70 by NM/22 for Bugzilla 607.
3402 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3403 increasing numbers of administrators be bitten by this.
3405 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3407 PP/03 Bugzilla 994: added openssl_options main configuration option.
3409 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3411 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3413 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3415 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3416 without a peer certificate, leading to a segfault because of an
3417 assumption that peers always have certificates. Be a little more
3418 paranoid. Problem reported by Martin Tscholak.
3420 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3421 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3422 NB: ClamAV planning to remove STREAM in "middle of 2010".
3423 CL also introduces -bmalware, various -d+acl logging additions and
3424 more caution in buffer sizes.
3426 PP/09 Implemented reverse_ip expansion operator.
3428 PP/10 Bugzilla 937: provide a "debug" ACL control.
3430 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3432 PP/12 Bugzilla 973: Implement --version.
3434 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3436 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3438 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3440 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3441 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3442 ignore trailing whitespace.
3444 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3446 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3449 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3450 Notification from Dr Andrew Aitchison.
3452 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3453 ExtendedDetectionInfo response format.
3454 Notification from John Horne.
3456 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3459 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3460 XSL and documented dependency on system catalogs, with examples of how
3463 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3466 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3467 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3468 configuration file which is writeable by the Exim user or group.
3470 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3471 of configuration files to cover files specified with the -C option if
3472 they are going to be used with root privileges, not just the default
3475 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3476 option (effectively making it always true).
3478 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3479 files to be used while preserving root privileges.
3481 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3482 that rogue child processes cannot use them.
3484 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3485 run-time user, instead of root.
3487 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3488 Exim run-time user without dropping privileges.
3490 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3491 result string, instead of calling string_vformat() twice with the same
3494 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3495 for other users. Others should always drop root privileges if they use
3496 -C on the command line, even for a whitelisted configure file.
3498 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3500 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3506 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3507 $data_path, and $header_path variables; fixed documentation bugs and
3510 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3511 exipick to access non-standard spools, including the "frozen" queue
3514 NM/01 Bugzilla 965: Support mysql stored procedures.
3515 Patch from Alain Williams
3517 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3519 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3520 Patch from Andreas Metzler
3522 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3523 Patch from Kirill Miazine
3525 NM/05 Bugzilla 671: Added umask to procmail example.
3527 JJ/03 installed exipick 20100323.0, fixing doc bug
3529 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3530 directory. Notification and patch from Dan Rosenberg.
3532 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3534 TK/02 Improve log output when DKIM signing operation fails.
3536 MH/01 Treat the transport option dkim_domain as a colon separated
3537 list, not as a single string, and sign the message with each element,
3538 omitting multiple occurences of the same signer.
3540 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3541 Bugzilla 985, 986. Patch by Simon Arlott
3543 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3544 Patch by Simon Arlott
3546 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3547 MBX locking. Notification from Dan Rosenberg.
3553 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3555 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3557 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3559 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3561 NM/04 Bugzilla 915: Flag broken perl installation during build.
3567 TK/01 Added patch by Johannes Berg that expands the main option
3568 "spamd_address" if it starts with a dollar sign.
3570 TK/02 Write list of recipients to X-Envelope-Sender header when building
3571 the mbox-format spool file for content scanning (suggested by Jakob
3574 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3575 (http://www.dcc-servers.net/) support via dccifd. Activated by
3576 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3578 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3579 by Mark Daniel Reidel <mr@df.eu>.
3581 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3582 When building exim an external PCRE library is now needed -
3583 PCRE is a system library on the majority of modern systems.
3584 See entry on PCRE_LIBS in EDITME file.
3586 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3587 conversation. Added nologin parameter to request.
3588 Patch contributed by Kirill Miazine.
3590 TF/01 Do not log submission mode rewrites if they do not change the address.
3592 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3594 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3595 log files in place. Contributed by Roberto Lima.
3597 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3599 TF/03 Bugzilla 615: When checking the local_parts router precondition
3600 after a local_part_suffix or local_part_prefix option, Exim now
3601 does not use the address's named list lookup cache, since this
3602 contains cached lookups for the whole local part.
3604 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3605 Robert Millan. Documentation is in experimental-spec.txt.
3607 TF/04 Bugzilla 668: Fix parallel build (make -j).
3609 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3611 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3612 Patch provided by Jan Srzednicki.
3614 TF/05 Leading white space used to be stripped from $spam_report which
3615 wrecked the formatting. Now it is preserved.
3617 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3618 that they are available at delivery time.
3620 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3622 TF/08 TLS error reporting now respects the incoming_interface and
3623 incoming_port log selectors.
3625 TF/09 Produce a more useful error message if an SMTP transport's hosts
3626 setting expands to an empty string.
3628 NM/06 Bugzilla 744: EXPN did not work under TLS.
3629 Patch provided by Phil Pennock.
3631 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3632 Patch provided by Richard Godbee.
3634 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3635 acl_smtp_notquit, added index entry.
3637 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3638 Patch provided by Eugene Bujak.
3640 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3641 accept(). Patch provided by Maxim Dounin.
3643 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3644 Patch provided by Phil Pennock.
3646 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3648 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3649 Patch provided by Brad "anomie" Jorsch.
3651 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3652 Patch provided by Dean Brooks.
3654 TK/05 Add native DKIM support (does not depend on external libraries).
3656 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3657 Patch provided by Graeme Fowler.
3659 NM/16 Bugzilla 851: Documentation example syntax fix.
3661 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3663 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3666 NM/19 Bugzilla 745: TLS version reporting.
3667 Patch provided by Phil Pennock.
3669 NM/20 Bugzilla 167: bool: condition support.
3670 Patch provided by Phil Pennock.
3672 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3673 clients. Patch provided by Phil Pennock.
3675 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3676 Patch provided by Brad "anomie" Jorsch.
3678 NM/23 Bugzilla 687: Fix misparses in eximstats.
3679 Patch provided by Heiko Schlittermann.
3681 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3682 Patch provided by Heiko Schlittermann.
3684 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3685 plus update to original patch.
3687 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3689 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3690 Patch provided by David Brownlee.
3692 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3694 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3696 NM/30 Bugzilla 888: TLS documentation bugfixes.
3698 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3700 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3701 Unlike the original bugzilla I have changed all shell scripts in src tree.
3703 NM/33 Bugzilla 898: Transport filter timeout fix.
3704 Patch by Todd Rinaldo.
3706 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3707 Patch by Serge Demonchaux.
3709 NM/35 Bugzilla 39: Base64 decode bug fixes.
3710 Patch by Jakob Hirsch.
3712 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3714 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3716 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3718 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3724 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3725 ALT-N's libdkim that I have put here:
3726 http://duncanthrax.net/exim-experimental/
3728 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3729 were called 'true' and it seems that C99 defines that as a reserved
3730 keyword to be used with 'bool' variable types. That means you could
3731 not include C99-style headers which use bools without triggering
3732 build errors in sieve.c.
3734 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3735 as mailq or other aliases. Changed the --help handling significantly
3736 to do whats expected. exim_usage() emits usage/help information.
3738 SC/01 Added the -bylocaldomain option to eximstats.
3740 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3742 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3744 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3750 PH/01 Another patch from the Sieve maintainer.
3752 PH/02 When an IPv6 address is converted to a string for single-key lookup
3753 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3754 dots are used instead of colons so that keys in lsearch files need not
3755 contain colons. This was done some time before quoting was made available
3756 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3757 (notated using the quote facility) so as to distinguish them from IPv4
3758 keys. This meant that lookups for IP addresses in host lists did not work
3759 for iplsearch lookups.
3761 This has been fixed by arranging for IPv6 addresses to be expressed with
3762 colons if the lookup type is iplsearch. This is not incompatible, because
3763 previously such lookups could never work.
3765 The situation is now rather anomalous, since one *can* have colons in
3766 ordinary lsearch keys. However, making the change in all cases is
3767 incompatible and would probably break a number of configurations.
3769 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3772 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3773 conversion specification without a maximum field width, thereby enabling
3774 a rogue spamd server to cause a buffer overflow. While nobody in their
3775 right mind would setup Exim to query an untrusted spamd server, an
3776 attacker that gains access to a server running spamd could potentially
3777 exploit this vulnerability to run arbitrary code as the Exim user.
3779 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3780 $primary_hostname instead of what libspf2 thinks the hosts name is.
3782 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3783 a directory entry by the name of the lookup key. Previously, if a
3784 symlink pointed to a non-existing file or a file in a directory that
3785 Exim lacked permissions to read, a lookup for a key matching that
3786 symlink would fail. Now it is enough that a matching directory entry
3787 exists, symlink or not. (Bugzilla 503.)
3789 PH/03 The body_linecount and body_zerocount variables are now exported in the
3792 PH/04 Added the $dnslist_matched variable.
3794 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3795 This means they are set thereafter only if the connection becomes
3798 PH/06 Added the client_condition to authenticators so that some can be skipped
3799 by clients under certain conditions.
3801 PH/07 The error message for a badly-placed control=no_multiline_responses left
3802 "_responses" off the end of the name.
3804 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3806 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3807 (without spaces) instead of just copying the configuration text.
3809 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3811 PH/11 Added $max_received_linelength.
3813 PH/12 Added +ignore_defer and +include_defer to host lists.
3815 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3816 way in which PCRE > 7.0 is built.
3818 PH/14 Implemented queue_only_load_latch.
3820 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3821 MAIL command. The effect was to mangle the value on 64-bit systems.
3823 PH/16 Another patch from the Sieve maintainer.
3825 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3827 PH/18 If a system quota error occurred while trying to create the file for
3828 a maildir delivery, the message "Mailbox is full" was not appended to the
3829 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3830 only to a quota excession during the actual writing of the file.
3832 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3833 characters?) which causes problems in log lines. The DN values are now
3834 passed through string_printing() before being added to log lines.
3836 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3837 and InterBase are left for another time.)
3839 PH/21 Added message_body_newlines option.
3841 PH/22 Guard against possible overflow in moan_check_errorcopy().
3843 PH/23 POSIX allows open() to be a macro; guard against that.
3845 PH/24 If the recipient of an error message contained an @ in the local part
3846 (suitably quoted, of course), incorrect values were put in $domain and
3847 $local_part during the evaluation of errors_copy.
3853 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3854 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3857 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3858 issue a MAIL command.
3860 PH/02 In an ACL statement such as
3862 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3864 if a client was not listed at all, or was listed with a value other than
3865 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3866 the condition was not true (as it should be), so access was not denied.
3867 The bug was that the ! inversion was incorrectly passed on to the second
3868 item. This has been fixed.
3870 PH/03 Added additional dnslists conditions == and =& which are different from
3871 = and & when the dns lookup returns more than one IP address.
3873 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3874 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3876 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3877 FSYNC, which compiles an option called disable_fsync that allows for
3878 bypassing fsync(). The documentation is heavily laced with warnings.
3880 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3882 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3883 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3884 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3885 including adding "make clean"; (3) Added -fPIC when compiling the test
3886 dynamically loaded module, to get rid of a warning.
3888 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3889 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3890 = 0s. The bug is otherwise harmless.
3892 PH/07 There was a bug in the dovecot authenticator such that the value of
3893 $auth1 could be overwritten, and so not correctly preserved, after a
3894 successful authentication. This usually meant that the value preserved by
3895 the server_setid option was incorrect.
3897 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3899 PH/09 Installed PCRE release 7.0.
3901 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3902 run for batched SMTP input. It is now run at the start of every message
3903 in the batch. While fixing this I discovered that the process information
3904 (output by running exiwhat) was not always getting set for -bs and -bS
3905 input. This is fixed, and it now also says "batched" for BSMTP.
3907 PH/11 Added control=no_pipelining.
3909 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3910 patch, slightly modified), and move the expansion of helo_data till after
3911 the connection is made in the smtp transport (so it can use these
3914 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3916 PH/14 Added log_selector = +pid.
3918 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3920 PH/16 Add ${if forany and ${if forall.
3922 PH/17 Added dsn_from option to vary the From: line in DSNs.
3924 PH/18 Flush SMTP output before performing a callout, unless control =
3925 no_callout_flush is set.
3927 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3928 was true (the default) a successful delivery failed to delete the retry
3929 item, thus causing premature timeout of the address. The bug is now
3932 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3934 PH/21 Long custom messages for fakedefer and fakereject are now split up
3935 into multiline responses in the same way that messages for "deny" and
3936 other ACL rejections are.
3938 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3939 with slight modification.
3941 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3942 draft, changing the syntax and factoring some duplicate code".
3944 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3945 for deliveries of the second and subsequent messages over the same SMTP
3948 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3949 ${reduce, with only minor "tidies".
3951 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3953 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3954 expansion side effects.
3956 PH/27 When a message times out after an over-quota error from an Exim-imposed
3957 quota, the bounce message says "mailbox is full". This message was not
3958 being given when it was a system quota that was exceeded. It now should
3961 MH/03 Made $recipients available in local_scan(). local_scan() already has
3962 better access to the recipient list through recipients_list[], but
3963 $recipients can be useful in postmaster-provided expansion strings.
3965 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3966 in the case of a MAIL command with additional options following the
3967 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3968 were accidentally chopped off.
3970 PH/29 SMTP synchronization checks are implemented when a command is read -
3971 there is a check that no more input is waiting when there shouldn't be
3972 any. However, for some commands, a delay in an ACL can mean that it is
3973 some time before the response is written. In this time, more input might
3974 arrive, invalidly. So now there are extra checks after an ACL has run for
3975 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3976 pipelining has not been advertised.
3978 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3980 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3981 returned in the SMTP response when smtp_return_error_details was false.
3982 This has been fixed.
3984 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3985 instead of the C I/O that was originally supplied, because problems were
3986 reported on Solaris.
3988 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3989 Exim which did not show up earlier: it was assuming that a call to
3990 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3991 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3992 was a macro that became an assignment, so it seemed to work. This has
3993 changed to a proper function call with a void return, hence the compile
3994 error. Exim's code has been fixed.
3996 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3999 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4001 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4003 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4004 because it turns out to be too big or there is a timeout) there is no
4005 "Completed" line in the log. When some messages of this type were
4006 selected by exigrep, they were listed as "not completed". Others were
4007 picked up by some special patterns. I have improved the selection
4008 criteria to be more general.
4010 PH/38 The host_find_failed option in the manualroute router can now be set
4011 to "ignore", to completely ignore a host whose IP address cannot be
4012 found. If all hosts are ignored, the behaviour is controlled by the new
4013 host_all_ignored option.
4015 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4016 homing or because of multiple MX records with /mx) generated more than
4017 one IP address, and the following item turned out to be the local host,
4018 all the secondary addresses of the first item were incorrectly removed
4019 from the list, along with the local host and any following hosts (which
4020 is what is supposed to happen).
4022 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4023 whoever called Exim into the -H file. In the case of the daemon it was
4024 behaving confusingly. When first started, it used values for whoever
4025 started the daemon, but after a SIGHUP it used the Exim user (because it
4026 calls itself on a restart). I have changed the code so that it now always
4029 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4030 message are rejected with the same error (e.g. no authentication or bad
4031 sender address), and a DATA command is nevertheless sent (as can happen
4032 with PIPELINING or a stupid MUA), the error message that was given to the
4033 RCPT commands is included in the rejection of the DATA command. This is
4034 intended to be helpful for MUAs that show only the final error to their
4037 PH/42 Another patch from the Sieve maintainer.
4039 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4040 Eximstats - Fixed some broken HTML links and added missing column headers
4042 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4043 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4045 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4047 PH/43 Yet another patch from the Sieve maintainer.
4049 PH/44 I found a way to check for a TCP/IP connection going away before sending
4050 the response to the final '.' that terminates a message, but only in the
4051 case where the client has not sent further data following the '.'
4052 (unfortunately, this is allowed). However, in many cases there won't be
4053 any further data because there won't be any more messages to send. A call
4054 to select() can be used: if it shows that the input is "ready", there is
4055 either input waiting, or the socket has been closed. An attempt to read
4056 the next input character can distinguish the two cases. Previously, Exim
4057 would have sent an OK response which the client would never have see.
4058 This could lead to message repetition. This fix should cure that, at
4059 least in a lot of common cases.
4061 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4062 advertised in response to EHLO.
4068 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4069 fixed by 4.65/MH/01 (is this a record?) are fixed:
4071 (i) An empty string was always treated as zero by the numeric comparison
4072 operators. This behaviour has been restored.
4074 (ii) It is documented that the numeric comparison operators always treat
4075 their arguments as decimal numbers. This was broken in that numbers
4076 starting with 0 were being interpreted as octal.
4078 While fixing these problems I realized that there was another issue that
4079 hadn't been noticed. Values of message_size_limit (both the global option
4080 and the transport option) were treated as octal if they started with 0.
4081 The documentation was vague. These values are now always treated as
4082 decimal, and I will make that clear in the documentation.
4088 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4089 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4092 MH/01 Don't check that the operands of numeric comparison operators are
4093 integers when their expansion is in "skipping" mode (fixes bug
4094 introduced by 4.64-PH/07).
4096 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4097 child addresses, Exim now panics and dies. Previously, because the count
4098 is held in a short int, deliveries were likely to be lost. As such a
4099 large number of recipients for a single message is ridiculous
4100 (performance will be very, very poor), I have chosen to impose a limit
4101 rather than extend the field.
4107 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4108 leftover -K file (the existence of which was triggered by #402).
4109 While we were at it, introduced process PID as part of the -K
4110 filename. This should rule out race conditions when creating
4113 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4114 processing considerably. Previous code took too long for large mails,
4115 triggering a timeout which in turn triggers #401.
4117 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4118 in the DK code in transports.c. sendfile() is not really portable,
4119 hence the _LINUX specificness.
4121 TF/01 In the add_headers option to the mail command in an Exim filter,
4122 there was a bug that Exim would claim a syntax error in any
4123 header after the first one which had an odd number of characters
4126 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4127 callout verification, Exim cached a "reject" for the entire domain. This
4128 is correct for most verifications, but it is not correct for a recipient
4129 verification with use_sender or use_postmaster set, because in that case
4130 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4131 case of MAIL FROM:<> rejection from other early rejections (e.g.
4132 rejection of HELO). When verifying a recipient using a non-null MAIL
4133 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4134 Whatever the result of the callout, the value of the domain cache is
4135 left unchanged (for any other kind of callout, getting as far as trying
4136 RCPT means that the domain itself is ok).
4138 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4141 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4142 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4143 session, the daemon ignores ECONNECT errors and logs others; it now
4144 ignores EPIPE as well.
4146 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4147 (quoted-printable decoding).
4149 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4150 later the small subsequent patch to fix an introduced bug.
4152 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4154 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4156 PH/08 An error is now given if message_size_limit is specified negative.
4158 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4159 to be given (somewhat) arbitrary names.
4161 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4164 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4165 miscellaneous code fixes
4167 PH/10 Added the log_reject_target ACL modifier to specify where to log
4170 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4171 hostname. This is wrong, because it relates to the incoming message (and
4172 probably the interface on which it is arriving) and not to the outgoing
4173 callout (which could be using a different interface). This has been
4174 changed to use the value of the helo_data option from the smtp transport
4175 instead - this is what is used when a message is actually being sent. If
4176 there is no remote transport (possible with a router that sets up host
4177 addresses), $smtp_active_hostname is used.
4179 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4180 tweaks were necessary in order to get it to work (see also 21 below):
4181 (a) The code assumed that strncpy() returns a negative number on buffer
4182 overflow, which isn't the case. Replaced with Exim's string_format()
4184 (b) There were several signed/unsigned issues. I just did the minimum
4185 hacking in of casts. There is scope for a larger refactoring.
4186 (c) The code used strcasecmp() which is not a standard C function.
4187 Replaced with Exim's strcmpic() function.
4188 (d) The code set only $1; it now sets $auth1 as well.
4189 (e) A simple test gave the error "authentication client didn't specify
4190 service in request". It would seem that Dovecot has changed its
4191 interface. Fortunately there's a specification; I followed it and
4192 changed what the client sends and it appears to be working now.
4194 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4197 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4198 address A is aliased to B and C, where B exists and C does not. Without
4199 -v the output is "A verified" because verification stops after a
4200 successful redirection if more than one address is generated. However,
4201 with -v the child addresses are also verified. Exim was outputting "A
4202 failed to verify" and then showing the successful verification for C,
4203 with its parentage. It now outputs "B failed to verify", showing B's
4204 parentage before showing the successful verification of C.
4206 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4207 look up a TXT record in a specific list after matching in a combined
4210 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4211 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4212 they consult the DNS. I had assumed they would set it the way they
4213 wanted; and indeed my experiments on Linux seem to show that in some
4214 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4215 To be on the safe side, however, I have now made the interface to
4216 host_find_byname() similar to host_find_bydns(), with an argument
4217 containing the DNS resolver options. The host_find_byname() function now
4218 sets these options at its start, just as host_find_bydns() does. The smtp
4219 transport options dns_qualify_single and dns_search_parents are passed to
4220 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4221 of host_find_byname() use the default settings of RES_DEFNAMES
4222 (qualify_single) but not RES_DNSRCH (search_parents).
4224 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4225 spool_read_header() do less string testing, by means of a preliminary
4226 switch on the second character of optional "-foo" lines. (This is
4227 overdue, caused by the large number of possibilities that now exist.
4228 Originally there were few.) While I was there, I also converted the
4229 str(n)cmp tests so they don't re-test the leading "-" and the first
4230 character, in the hope this might squeeze out yet more improvement.
4232 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4233 flag allowing group syntax was set by the header_syntax check but not
4234 turned off, possible causing trouble later; (2) The flag was not being
4235 set at all for the header_verify test, causing "group"-style headers to
4236 be rejected. I have now set it in this case, and also caused header_
4237 verify to ignore an empty address taken from a group. While doing this, I
4238 came across some other cases where the code for allowing group syntax
4239 while scanning a header line wasn't quite right (mostly, not resetting
4240 the flag correctly in the right place). These bugs could have caused
4241 trouble for malformed header lines. I hope it is now all correct.
4243 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4244 with the "reply" argument non-NULL. The code, however (which originally
4245 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4246 but it didn't always do it. This confused somebody who was copying the
4247 code for some other use. I have removed all the tests.
4249 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4250 feature that was used to support insecure browsers during the U.S. crypto
4251 embargo. It requires special client support, and Exim is probably the
4252 only MTA that supported it -- and would never use it because real RSA is
4253 always available. This code has been removed, because it had the bad
4254 effect of slowing Exim down by computing (never used) parameters for the
4255 RSA_EXPORT functionality.
4257 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4258 authenticator to fail if there's a tab character in the incoming data
4259 (there should never be unless someone is messing about, as it's supposed
4260 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4261 if the connection is using TLS or if the remote IP is the same as the
4262 local IP, and the "valid-client-cert option" if a client certificate has
4265 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4266 authenticators. This can be used for authorization after authentication
4267 succeeds. (In the case of plaintext, it servers for both authentication
4270 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4271 if any retry times were supplied.
4273 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4274 connection, where there is no client IP address to check. In this
4275 situation, the verify now always succeeds.
4277 PH/25 Applied John Jetmore's -Mset patch.
4279 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4281 PH/27 In a string expansion for a processed (not raw) header when multiple
4282 headers of the same name were present, leading whitespace was being
4283 removed from all of them, but trailing whitespace was being removed only
4284 from the last one. Now trailing whitespace is removed from each header
4285 before concatenation. Completely empty headers in a concatenation (as
4286 before) are ignored.
4288 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4289 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4291 PH/29 [Removed. This was a change that I later backed out, and forgot to
4292 correct the ChangeLog entry (that I had efficiently created) before
4293 committing the later change.]
4295 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4296 address errors (4xx response to RCPT) over the same connection as other
4297 messages routed to the same hosts. Such deliveries are always "forced",
4298 so retry times are not inspected. This resulted in far too many retries
4299 for the affected addresses. The effect occurred only when there were more
4300 hosts than the hosts_max_try setting in the smtp transport when it had
4301 the 4xx errors. Those hosts that it had tried were not added to the list
4302 of hosts for which the message was waiting, so if all were tried, there
4303 was no problem. Two fixes have been applied:
4305 (i) If there are any address or message errors in an SMTP delivery, none
4306 of the hosts (tried or untried) are now added to the list of hosts
4307 for which the message is waiting, so the message should not be a
4308 candidate for sending over the same connection that was used for a
4309 successful delivery of some other message. This seems entirely
4310 reasonable: after all the message is NOT "waiting for some host".
4311 This is so "obvious" that I'm not sure why it wasn't done
4312 previously. Hope I haven't missed anything, but it can't do any
4313 harm, as the worst effect is to miss an optimization.
4315 (ii) If, despite (i), such a delivery is accidentally attempted, the
4316 routing retry time is respected, so at least it doesn't keep
4317 hammering the server.
4319 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4320 in ${readsocket because some servers need this prod.
4322 PH/32 Added some extra debug output when updating a wait-xxx database.
4324 PH/33 The hint "could be header name not terminated by colon", which has been
4325 given for certain expansion errors for a long time, was not being given
4326 for the ${if def:h_colon_omitted{... case.
4328 PH/34 The spec says: "With one important exception, whenever a domain list is
4329 being scanned, $domain contains the subject domain." There was at least
4330 one case where this was not true.
4332 PH/35 The error "getsockname() failed: connection reset by peer" was being
4333 written to the panic log as well as the main log, but it isn't really
4334 panic-worthy as it just means the connection died rather early on. I have
4335 removed the panic log writing for the ECONNRESET error when getsockname()
4338 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4339 runs only) independently of the message's sender address. This meant
4340 that, if the 4xx error was in fact related to the sender, a different
4341 message to the same recipient with a different sender could confuse
4342 things. In particular, this can happen when sending to a greylisting
4343 server, but other circumstances could also provoke similar problems.
4344 I have changed the default so that the retry time for these errors is now
4345 based a combination of the sender and recipient addresses. This change
4346 can be overridden by setting address_retry_include_sender=false in the
4349 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4350 remote server are returned as part of bounce messages. This was not
4351 happening for LMTP over a pipe (the lmtp transport), but now it is the
4352 same for both kinds of LMTP.
4354 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4355 in header lines that were in fact CNAMEs. This is no longer the case.
4357 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4358 and queue runs started by the daemon processed all messages. This has
4359 been fixed so that -R and -S can now usefully be given with -q<time>.
4361 PH/40 Import PCRE release 6.7 (fixes some bugs).
4363 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4365 PH/42 Give an error if -q is specified more than once.
4367 PH/43 Renamed the variables $interface_address and $interface_port as
4368 $received_ip_address and $received_port, to make it clear that these
4369 values apply to message reception, and not to the outgoing interface when
4370 a message is delivered. (The old names remain recognized, of course.)
4372 PH/44 There was no timeout on the connect() call when using a Unix domain
4373 socket in the ${readsocket expansion. There now is.
4375 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4376 be meaningful with "accept".
4378 SC/01 Eximstats V1.43
4379 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4381 SC/02 Eximstats V1.44
4382 Use a glob alias rather than an array ref in the generated
4383 parser. This improves both readability and performance.
4385 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4386 Collect SpamAssassin and rejection statistics.
4387 Don't display local sender or destination tables unless
4388 there is data to show.
4389 Added average volumes into the top table text output.
4391 SC/04 Eximstats V1.46
4392 Collect data on the number of addresses (recipients)
4393 as well as the number of messages.
4395 SC/05 Eximstats V1.47
4396 Added 'Message too big' to the list of mail rejection
4397 reasons (thanks to Marco Gaiarin).
4399 SC/06 Eximstats V1.48
4400 Mainlog lines which have GMT offsets and are too short to
4401 have a flag are now skipped.
4403 SC/07 Eximstats V1.49 (Alain Williams)
4404 Added the -emptyok flag.
4406 SC/08 Eximstats V1.50
4407 Fixes for obtaining the IP address from reject messages.
4409 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4410 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4411 whitespace changes from 4.64-PH/27
4413 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4416 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4417 are found, allow negative numbers in numeric criteria)
4419 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4421 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4422 to match changes made in 4.64-PH/43
4424 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4426 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4427 "domain" is actually an IP address, and force a failure. This locks out
4428 those revolvers/nameservers that support "A-for-A" lookups, in
4429 contravention of the specifications.
4431 PH/48 When a host name was looked up from an IP address, and the subsequent
4432 forward lookup of the name timed out, the host name was left in
4433 $sender_host_name, contrary to the specification.
4435 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4436 restricted to single-key lookups, Exim was not diagnosing an error if
4437 * or *@ was used with a query-style lookup.
4439 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4441 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4442 long ago, but noone interested enough thought of it. Let's just say that
4443 the "1.1" means that there are some new functions that weren't there at
4444 some point in the past.
4446 PH/51 Error processing for expansion failure of helo_data from an smtp
4447 transport during callout processing was broken.
4449 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4450 tested/used via the -bh/-bhc/-bs options.
4452 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4453 bug, fixed in subsequent PCRE releases).
4455 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4456 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4458 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4464 SC/01 Use a glob alias rather than an array ref in eximstats generated
4465 parser. This improves both readability and performance.
4467 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4468 Don't display local sender or destination tables in eximstats unless
4469 there is data to show.
4470 Added average volumes into the eximstats top table text output.
4472 SC/03 Collect data on the number of addresses (recipients) as well
4473 as the number of messages in eximstats.
4475 TF/01 Correct an error in the documentation for the redirect router. Exim
4476 does (usually) call initgroups() when daemonizing.
4478 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4479 with consistent privilege compared to when running as a daemon.
4481 TF/03 Note in the spec that $authenticated_id is not set for local
4482 submissions from trusted users.
4484 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4485 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4487 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4488 by adding some example configuration directives to the default
4489 configuration file. A little bit of work is required to uncomment the
4490 directives and define how usernames and passwords are checked, but
4491 there is now a framework to start from.
4493 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4494 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4495 without this. I don't know how relevant this is to other LDAP libraries.
4497 PH/02 Add the verb name to the "unknown ACL verb" error.
4499 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4501 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4503 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4504 directory not expanded when it should be if an expanded home directory
4505 was set for the address (which is overridden by the transport).
4507 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4510 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4511 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4512 because it is too late at that time, and has no effect.
4514 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4515 security issue with \' (bugzilla #107). I could not use the
4516 PQescapeStringConn() function, because it needs a PGconn value as one of
4519 PH/08 When testing addresses using -bt, indicate those final addresses that
4520 are duplicates that would not cause an additional delivery. At least one
4521 person was confused, thinking that -bt output corresponded to deliveries.
4522 (Suppressing duplicates isn't a good idea as you lose the information
4523 about possibly different redirections that led to the duplicates.)
4525 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4526 systems where poll() doesn't work, in particular OS X.
4528 PH/10 Added more information to debugging output for retry time not reached.
4530 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4531 operations in malware.c.
4533 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4536 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4537 syslog_duplication set false, log lines that would normally be written
4538 both the the main log and to the reject log were not written to syslog at
4541 PH/14 In the default configuration, change the use of "message" in ACL warn
4542 statements to "add_header".
4544 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4545 not followed by a command (e.g. "seen endif").
4547 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4548 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4551 PH/17 Added extra conditions to the default value of delay_warning_condition
4555 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4556 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4557 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4560 The Auto-Submitted: and various List- headers are standardised, whereas I
4561 don't think Precedence: ever was.
4563 PH/18 Refactored debugging code in route_finduser() to show more information,
4564 in particular, the error code if getpwnam() issues one.
4566 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4567 This is apparently needed in addition to the PH/07 change above to avoid
4568 any possible encoding problems.
4570 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4571 but not after initializing Perl.
4573 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4574 output them only if debugging. By default they are written stderr,
4575 apparently, which is not desirable.
4577 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4580 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4583 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4585 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4586 authenticated or an ident call has been made. Suppress the default
4587 values for $authenticated_id and $authenticated_sender (but permit -oMai
4588 and -oMas) when testing with -bh.
4590 PH/24 Re-jigged the order of the tests in the default configuration so that the
4591 tests for valid domains and recipients precede the DNS black list and CSA
4592 tests, on the grounds that those ones are more expensive.
4594 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4595 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4596 This bug exists in every version of Exim that I still have, right back to
4599 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4600 However, an attempt to turn on case-sensitivity in a regex key by
4601 including (?-i) didn't work because the subject string was already
4602 lowercased, and the effects were non-intuitive. It turns out that a
4603 one-line patch can be used to allow (?-i) to work as expected.
4609 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4610 other effects) broke the use of negated acl sub-conditions.
4612 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4615 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4616 "Deny" causes Exim to reject the incoming connection with a 554 error.
4617 Unfortunately, if there is a major crisis, such as a disk failure,
4618 tcp-wrappers gives "deny", whereas what one would like would be some
4619 kind of temporary error. A kludge has been added to help with this.
4620 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4621 554 error is used if errno is still zero or contains ENOENT (which occurs
4622 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4625 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4627 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4628 errors. Otherwise a message that provokes a temporary error (when other
4629 messages do not) can cause a whole host to time out.
4631 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4632 addresses were routed directly to files or pipes from a redirect router.
4633 File deliveries just didn't batch; pipe deliveries might have suffered
4636 PH/06 A failure to get a lock for a hints database would erroneously always say
4637 "Failed to get write lock", even when it was really a read lock.
4639 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4640 of 0600. This has been changed to use the value of the lockfile_mode
4641 option (which defaults to 0600).
4643 PH/08 Applied small patch from the Sieve maintainer.
4645 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4646 folder from quota calculations, a direct delivery into this folder messed
4647 up the contents of the maildirsize file. This was because the regex was
4648 used only to exclude .Trash (or whatever) when the size of the mailbox
4649 was calculated. There was no check that a delivery was happening into an
4650 excluded directory. This bug has been fixed by ignoring all quota
4651 processing for deliveries into excluded directories.
4653 PH/10 Added the maildirfolder_create_regex option to appendfile.
4659 PH/01 The code for finding all the local interface addresses on a FreeBSD
4660 system running IPv6 was broken. This may well have applied to all BSD
4661 systems, as well as to others that have similar system calls. The broken
4662 code found IPv4 interfaces correctly, but gave incorrect values for the
4663 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4664 that it would not match correctly against @[] and not recognize the IPv6
4667 PH/02 The ipliteral router was not recognizing addresses of the form user@
4668 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4670 PH/03 Added disable_ipv6.
4672 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4673 decoded form, because it is most often used to construct To: headers
4674 lines in autoreplies, and the decoded form may well be syntactically
4675 invalid. However, $reply_address has leading white space removed, and all
4676 newlines turned into spaces so that the autoreply transport does not
4679 PH/05 If group was specified without a user on a router, and no group or user
4680 was specified on a transport, the group from the router was ignored.
4682 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4683 for visible compile-time settings that can be used to change these
4684 numbers, for those that want even more. Backwards compatibility with old
4685 spool files has been maintained. However, going back to a previous Exim
4686 release will lost any variables that are in spool files.
4688 PH/07 Two small changes when running in the test harness: increase delay when
4689 passing a TCP/IP connection to a new process, in case the original
4690 process has to generate a bounce, and remove special handling of
4691 127.0.0.2 (sic), which is no longer necessary.
4693 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4694 be the same on different OS.
4696 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4699 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4700 whether --show-vars was specified or not
4702 JJ/02 exipick: Added support for new ACL variable spool format introduced
4705 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4706 syntactically invalid From: or Reply-to: line, and a filter used this to
4707 generate an autoreply, and therefore failed to obtain an address for the
4708 autoreply, Exim could try to deliver to a non-existent relative file
4709 name, causing unrelated and misleading errors. What now happens is that
4710 it logs this as a hard delivery error, but does not attempt to create a
4713 PH/11 The exinext utility has a -C option for testing purposes, but although
4714 the given file was scanned by exinext itself; it wasn't being passed on
4715 when Exim was called.
4717 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4718 an end-of-file indication when reading a command response.
4720 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4721 compiled. In many other places in Exim, IPv6 addresses are always
4722 recognized, so I have changed this. It also means that IPv4 domain
4723 literals of the form [IPV4:n.n.n.n] are now always recognized.
4725 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4726 used if the router is not running as root, for example, when verifying at
4727 ACL time, or when using -bh. The debugging output from this situation was
4728 non-existent - all you got was a failure to exec. I have made two
4731 (a) Failures to set uid/gid, the current directory, or a process leader
4732 in a subprocess such as that created by queryprogram now generate
4733 suitable debugging output when -d is set.
4735 (b) The queryprogram router detects when it is not running as root,
4736 outputs suitable debugging information if -d is set, and then runs
4737 the subprocess without attempting to change uid/gid.
4739 PH/15 Minor change to Makefile for building test_host (undocumented testing
4742 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4743 additional section of a DNS packet that returns MX or SRV records.
4744 Instead, it always explicitly searches for A/AAAA records. This avoids
4745 major problems that occur when a DNS server includes only records of one
4746 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4747 fixed another bug: if SRV records were looked up and the corresponding
4748 address records were *not* found in the additional section, the port
4749 values from the SRV records were lost.
4751 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4752 using the correct key (the original address) when searching the retry
4753 rules in order to find which one to use for generating the retry hint.
4755 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4756 adding the default one. Similarly, if it contains a Reply-To: header, the
4757 errors_reply_to option, if set, is not used.
4759 PH/19 When calculating a retry time, Exim used to measure the "time since
4760 failure" by looking at the "first failed" field in the retry record. Now
4761 it does not use this if it is later than than the arrival time of the
4762 message. Instead it uses the arrival time. This makes for better
4763 behaviour in cases where some deliveries succeed, thus re-setting the
4764 "first failed" field. An example is a quota failure for a huge message
4765 when small messages continue to be delivered. Without this change, the
4766 "time since failure" will always be short, possible causing more frequent
4767 delivery attempts for the huge message than are intended.
4768 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4770 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4771 $1, $2, $3) because the numerical variables can be reset during some
4772 expansion items (e.g. "match"), thereby losing the authentication data.
4774 PH/21 Make -bV show the size of off_t variables so that the test suite can
4775 decide whether to run tests for quotas > 2G.
4777 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4778 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4779 filecount value is greater than 2G or if a quota value is greater than 2G
4780 on a system where the size of off_t is not greater than 4, a panic error
4783 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4784 never match. The debug and -bh output now contains an explicit error
4785 message indicating a malformed IPv4 address or mask.
4787 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4788 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4789 PH/23 above applies.
4791 PH/25 Do not write to syslog when running in the test harness. The only
4792 occasion when this arises is a failure to open the main or panic logs
4793 (for which there is an explicit test).
4795 PH/26 Added the /no_tell option to "control=freeze".
4797 PH/27 If a host name lookup failed very early in a connection, for example, if
4798 the IP address matched host_lookup and the reverse lookup yielded a name
4799 that did not have a forward lookup, an error message of the form "no IP
4800 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4801 could be logged. Now it outputs the IP address instead of "NULL".
4803 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4804 allows the sender and the authenticated sender to be set when
4805 submitting a message from within Exim. Since child_open_exim() is
4806 documented for local_scan(), the new function should be too.
4808 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4809 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4810 results in an empty string is now treated as unset.
4812 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4814 PH/31 Added sender_verify_fail logging option.
4816 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4817 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4818 tidied the source and removed it altogether.
4820 PH/33 When a queue run was abandoned because the load average was too high, a
4821 log line was always written; now it is written only if the queue_run log
4822 selector is set. In addition, the log line for abandonment now contains
4823 information about the queue run such as the pid. This is always present
4824 in "start" and "stop" lines but was omitted from the "abandon" line.
4826 PH/34 Omit spaces between a header name and the colon in the error message that
4827 is given when verify = headers_syntax fails (if there are lots of them,
4828 the message gets confusing).
4830 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4831 names, as there are now some PTR records that contain slashes. This check
4832 is only to protect against broken name servers that fall over on strange
4833 characters, so the fact that it applies to all lookups doesn't matter.
4835 PH/36 Now that the new test suite is complete, we can remove some of the
4836 special code in Exim that was needed for the old test suite. For example,
4837 sorting DNS records because real resolvers return them in an arbitrary
4838 order. The new test suite's fake resolver always returns records in the
4841 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4842 bounces) except when queue_only is set, to avoid logging races between
4843 the different processes.
4845 PH/38 Panic-die if .include specifies a non-absolute path.
4847 PH/39 A tweak to the "H" retry rule from its user.
4849 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4850 a label. They prevented compilation on older perls.
4852 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4853 a warning to be raised on newish perls.
4855 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4856 on queue. Changes to match documented behaviour of showing count of
4857 messages matching specified criteria.
4859 PH/40 Changed the default ident timeout from 30s to 5s.
4861 PH/41 Added support for the use of login_cap features, on those BSD systems
4862 that have them, for controlling the resources used by pipe deliveries.
4864 PH/42 The content-scanning code uses fopen() to create files in which to put
4865 message data. Previously it was not paying any attention to the mode of
4866 the files. Exim runs with umask(0) because the rest of the code creates
4867 files with open(), and sets the required mode explicitly. Thus, these
4868 files were ending up world-writeable. This was not a big issue, because,
4869 being within the spool directory, they were not world-accessible. I have
4870 created a function called modefopen, which takes an additional mode
4871 argument. It sets umask(777), creates the file, chmods it to the required
4872 mode, then resets the umask. All the relevant calls to fopen() in the
4873 content scanning code have been changed to use this function.
4875 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4876 to 24 hours. This avoids potential overflow problems when processing G
4877 and H retry rules. I suspect nobody ever tinkers with this value.
4879 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4881 PH/45 When the plaintext authenticator is running as a client, the server's
4882 challenges are checked to ensure they are valid base64 strings. By
4883 default, the authentication attempt is cancelled if an invalid string is
4884 received. Setting client_ignore_invalid_base64 true ignores these errors.
4885 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4886 they are received. Thus, the responses can be made to depend on the
4887 challenges. If an invalid string is ignored, an empty string is placed in
4890 PH/46 Messages that are created by the autoreply transport now contains a
4891 References: header, in accordance with RFCs 2822 and 3834.
4893 PH/47 Added authenticated_sender_force to the smtp transport.
4895 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4897 PH/49 Installed latest patch from the Sieve maintainer.
4899 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4900 was also set, the appendfile transport was unnecessarily scanning a
4901 directory of message files (e.g. for maildir delivery) to find the count
4902 of files (along with the size), even though it did not need this
4903 information. It now does the scan only if it needs to find either the
4904 size of the count of files.
4906 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4908 PH/52 Two bugs concerned with error handling when the smtp transport is
4911 (i) Exim was not creating retry information for temporary errors given
4912 for individual recipients after the DATA command when the smtp transport
4913 was used in LMTP mode. This meant that they could be retried too
4914 frequently, and not timed out correctly.
4916 (ii) Exim was setting the flag that allows error details to be returned
4917 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4918 recipients that were returned after the DATA command.
4920 PH/53 This is related to PH/52, but is more general: for any failing address,
4921 when detailed error information was permitted to be returned to the
4922 sender, but the error was temporary, then after the final timeout, only
4923 "retry timeout exceeded" was returned. Now it returns the full error as
4924 well as "retry timeout exceeded".
4926 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4927 do this, and (what is worse) MTAs that accept it.
4929 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4930 will now be deprecated.
4932 PH/56 New os.c-cygwin from the Cygwin maintainer.
4934 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4935 formats (previously only available in exim formats via -bpr, -bpru,
4936 and -bpra. Now also available in native and exiqgrep formats)
4938 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4939 with very large, slow to parse queues
4941 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4943 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4945 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4946 responses to authentication challenges, though it was showing the
4947 challenges; (ii) I've removed the CR characters from the debug output for
4950 PH/58 Allow for the insertion of a newline as well as a space when a string
4951 is turned into more than one encoded-word during RFC 2047 encoding. The
4952 Sieve code now uses this.
4954 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4955 data_4xx, lost_connection, tls_required.
4957 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4958 message was being sent as an SMTP response.
4960 PH/61 Add -l and -k options to exicyclog.
4962 PH/62 When verifying, if an address was redirected to one new address, so that
4963 verification continued, and the new address failed or deferred after
4964 having set something in $address_data, the value of $address_data was not
4965 passed back to the ACL. This was different to the case when no
4966 redirection occurred. The value is now passed back in both cases.
4968 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4969 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4970 use login_cap.h, so on its own it isn't the distinguishing feature. The
4971 new name refers directly to the setclassresources() function.
4973 PH/65 Added configuration files for NetBSD3.
4975 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4977 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4978 is preferred over IPv4.
4980 PH/68 The bounce_return_message and bounce_return_body options were not being
4981 honoured for bounces generated during the reception of non-SMTP messages.
4982 In particular, this applied to messages rejected by the ACL. This bug has
4983 been fixed. However, if bounce_return_message is true and bounce_return_
4984 body is false, the headers that are returned for a non-SMTP message
4985 include only those that have been read before the error was detected.
4986 (In the case of an ACL rejection, they have all been read.)
4988 PH/69 The HTML version of the specification is now built in a directory called
4989 spec_html instead of spec.html, because the latter looks like a path with
4990 a MIME-type, and this confuses some software.
4992 PH/70 Catch two compiler warnings in sieve.c.
4994 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4995 function verify_get_ident() calls ip_connect() to connect a socket, but
4996 if the "connect()" function timed out, ip_connect() used to close the
4997 socket. However, verify_get_ident() also closes the socket later, and in
4998 between Exim writes to the log, which may get opened at this point. When
4999 the socket was closed in ip_connect(), the log could get the same file
5000 descriptor number as the socket. This naturally causes chaos. The fix is
5001 not to close the socket in ip_connect(); the socket should be closed by
5002 the function that creates it. There was only one place in the code where
5003 this was missing, in the iplookup router, which I don't think anybody now
5004 uses, but I've fixed it anyway.
5006 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5007 well as to direct DNS lookups. Otherwise the handling of names in host
5008 lists is inconsistent and therefore confusing.
5014 PH/01 Two changes to the default runtime configuration:
5016 (1) Move the checks for relay_from_hosts and authenticated clients from
5017 after to before the (commented out) DNS black list checks.
5019 (2) Add control=submission to the relay_from_hosts and authenticated
5020 clients checks, on the grounds that messages accepted by these
5021 statements are most likely to be submissions.
5023 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5025 (1) Generate an error if the third argument for the ${prvs expansion is
5028 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5031 (3) Reset the variables that are obtained from the first argument of
5032 ${prvscheck and used in the second argument before leaving the code,
5033 because their memory is reclaimed, so using them afterwards may do
5036 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5037 one (it's much easier than Tom thought :-).
5039 (5) Because of (4), we can now allow for the use of $prvscheck_result
5040 inside the third argument.
5042 PH/03 For some reason, the default setting of PATH when running a command from
5043 a pipe transport was just "/usr/bin". I have changed it to
5046 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5047 anything to be listed in the output from -bV.
5049 PH/05 When a filter generated an autoreply, the entire To: header line was
5050 quoted in the delivery log line, like this:
5052 => >A.N.Other <ano@some.domain> <original@ddress> ...
5054 This has been changed so that it extracts the operative address. There
5055 may be more than one such address. If so, they are comma-separated, like
5058 => >ano@some.domain,ona@other.domain <original@ddress> ...
5060 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5061 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5062 not being looked up in the rDNS to get a host name, Exim was showing the
5063 IP address twice in Received: lines, even though the IP addresses were
5064 identical. For example:
5066 Received: from [1.2.3.4] (helo=[1.2.3.4])
5068 However, if the real host name was known, it was omitting the HELO data
5069 if it matched the actual IP address. This has been tidied up so that it
5070 doesn't show the same IP address twice.
5072 PH/07 When both +timestamp and +memory debugging was on, the value given by
5073 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5074 called by the debug printing, thereby overwriting the timestamp buffer.
5075 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5077 PH/08 When the original message was included in an autoreply transport, it
5078 always said "this is a copy of the message, including all the headers",
5079 even if body_only or headers_only was set. It now gives an appropriate
5082 PH/09 Applied a patch from the Sieve maintainer which:
5084 o fixes some comments
5085 o adds the (disabled) notify extension core
5086 o adds some debug output for the result of if/elsif tests
5087 o points to the current vacation draft in the documentation
5088 and documents the missing references header update
5092 o fixes a bug in processing the envelope test (when testing
5093 multiple envelope elements, the last element determined the
5096 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5097 Electronic Mail") by including:
5099 Auto-submitted: auto-generated
5101 in the messages that it generates (bounce messages and others, such as
5102 warnings). In the case of bounce messages for non-SMTP messages, there was
5103 also a typo: it was using "Auto_submitted" (underscore instead of
5104 hyphen). Since every message generated by Exim is necessarily in response
5105 to another message, thes have all been changed to:
5107 Auto-Submitted: auto-replied
5109 in accordance with these statements in the RFC:
5111 The auto-replied keyword:
5113 - SHOULD be used on messages sent in direct response to another
5114 message by an automatic process,
5116 - MUST NOT be used on manually-generated messages,
5118 - MAY be used on Delivery Status Notifications (DSNs) and Message
5119 Disposition Notifications (MDNs),
5121 - MUST NOT be used on messages generated by automatic or periodic
5122 processes, except for messages which are automatic responses to
5125 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5126 to the default Received: header definition.
5128 PH/12 Added log selector acl_warn_skipped (default on).
5130 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5131 variables because (a) they are in the wrong storage pool and (b) even if
5132 they were copied, it wouldn't work properly because of the caching.
5134 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5135 checking when decoding. Apparently there are clients that generate
5136 overlong encoded strings. Why am I not surprised?
5138 PH/15 If the first argument of "${if match_address" was not empty, but did not
5139 contain an "@" character, Exim crashed. Now it writes a panic log message
5140 and treats the condition as false.
5142 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5144 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5145 extension "envelope-auth". The code is finished and in agreement with
5146 other implementations, but there is no documentation so far and in fact,
5147 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5148 not changing the active code.
5150 Print executed "if" and "elsif" statements when debugging is used. This
5151 helps a great deal to understand what a filter does.
5153 Document more things not specified clearly in RFC3028. I had all this
5154 sorted out, when out of a sudden new issues came to my mind. Oops."
5156 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5159 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5160 canonical form (as documented). However, after a host name lookup from
5161 the IP address, check_host() was doing a simple string comparison with
5162 addresses acquired from the DNS when checking that the found name did
5163 have the original IP as one of its addresses. Since any found IPv6
5164 addresses are likely to be in abbreviated form, the comparison could
5165 fail. Luckily, there already exists a function for doing the comparison
5166 by converting both addresses to binary, so now that is used instead of
5167 the text comparison.
5169 PH/20 There was another similar case to PH/19, when a complete host name was
5170 given in a host list; looking up its IP address could give an abbreviated
5171 form, whereas the current host's name might or might not be abbreviated.
5172 The same fix has been applied.
5178 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5179 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5182 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5183 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5185 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5187 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5188 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5189 Also, for "Auto-Submitted", treat anything other than "no" as
5190 non-personal, in accordance with RFC 3834. (Previously it treated
5191 anything starting "auto-" as non-personal.)
5193 TF/01 The control=submission/name=... option had a problem with syntax
5194 errors if the name included a slash character. The /name= option
5195 now slurps the rest of the string, so it can include any characters
5196 but it must come last in the list of options (after /sender_retain
5199 PH/05 Some modifications to the interface to the fake nameserver for the new
5207 TK/01 Added the "success_on_redirect" address verification option. See
5208 NewStuff for rationale and an example.
5210 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5212 PH/02 Patch to exigrep to allow it to work on syslog lines.
5214 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5215 fread() to skip over the body file's header line, because in Cygwin the
5216 header line is locked and is inaccessible.
5218 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5219 co-exist for some time) to make it clear that it is the Exim ID that is
5220 referenced, not the Message-ID: header line.
5222 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5223 string_format() function, because snprintf() does not exist on all
5226 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5227 ${stat: expansion item.
5229 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5230 protocol synchronization error", to keep the pedants happy.
5232 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5233 well as for IRIX systems, when gcc is being used. See the host.c source
5236 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5238 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5241 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5242 transport and to the smtp transport in LMTP mode.
5244 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5246 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5247 run a filter in a subprocess. This could lead to confusion in subsequent
5248 lookups in the parent process. There should also be a search_tidyup() at
5249 the end of the subprocess.
5251 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5252 only if the host matched helo_try_verify_hosts, which caused the
5253 verification to occur when the EHLO/HELO command was issued. The ACL just
5254 tested the remembered result. Now, if a previous verification attempt has
5255 not happened, "verify = helo" does it there and then.
5257 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5259 TK/03 Fix log output including CR from clamd.
5261 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5262 exist provoked a memory error which could cause a segfault.
5264 PH/15 Installed PCRE 6.2
5266 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5268 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5269 of the problem. Specifically, suggested +O2 rather than +O1 for the
5272 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5274 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5275 record, the port was not correctly logged when the outgoing_port log
5276 selector was set (it logged the transort's default port).
5278 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5279 fallback_hosts, and "hosts" in the smtp transport.
5281 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5282 host errors such as "Connection refused".
5284 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5285 authentication with radiusclient 0.4.9:
5287 - Error returned from rc_read_config was caught wrongly
5288 - Username/password not passed on to radius server due to wrong length.
5290 The presumption is that some radiusclient API changes for 4.51/PH/17
5291 were not taken care of correctly. The code is still untested by me (my
5292 Linux distribution still has 0.3.2 of radiusclient), but it was
5293 contributed by a Radius user.
5295 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5296 expanding the "port" option of the smtp transport.
5298 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5299 while reading a MIME header. Thanks to Tom Hughes for a patch.
5301 PH/24 Include config.h inside local_scan.h so that configuration settings are
5304 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5305 that in an ACL for RCPT (for example), you can examine exactly what was
5308 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5309 commands, but it was not correctly comparing the address with the actual
5310 client host address. Thus, it would show the EHLO address in Received:
5311 header lines when this was not necessary.
5313 PH/27 Added the % operator to ${eval:}.
5315 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5316 it should be ignoring failures (because with -C, for example, it has lost
5317 privilege). It wasn't ignoring creation failures other than "already
5320 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5323 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5324 because an input line was too long, either on its own, or by virtue of
5325 too many continuations, the temporary file was not being removed, and the
5326 return code was incorrect.
5328 PH/31 Missing "BOOL" in function definition in filtertest.c.
5330 PH/32 Applied Sieve patches from the maintainer.
5332 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5334 PH/33 Added "verify = not_blind".
5336 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5337 Local/Makefile (with some defaults set). These are used in built scripts
5338 such as exicyclog, but they have never been used in the exim_install
5339 script (though there are many overriding facilities there). I have
5340 arranged that the exim_install script now takes note of these two
5343 PH/35 Installed configuration files for Dragonfly.
5345 PH/36 When a locally submitted message by a trusted user did not contain a
5346 From: header, and the sender address was obtained from -f or from an SMTP
5347 MAIL command, and the trusted user did not use -F to supply a sender
5348 name, $originator_name was incorrectly used when constructing a From:
5349 header. Furthermore, $originator_name was used for submission mode
5350 messages from external hosts without From: headers in a similar way,
5351 which is clearly wrong.
5353 PH/37 Added control=suppress_local_fixups.
5355 PH/38 When log_selector = +received_sender was set, and the addition of the
5356 sender made the log line's construction buffer exactly full, or one byte
5357 less than full, an overflow happened when the terminating "\n" was
5360 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5361 when the result of a list match is failure because a DNS lookup failed.
5363 PH/40 RM_COMMAND is now used in the building process.
5365 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5366 the "build-* directories that it finds.
5368 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5369 address in a domain literal was a prefix of an interface address.
5371 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5372 when verifying a sender address, unless rewrite_headers is false.
5374 PH/44 Wrote a long comment about why errors_to addresses are verified as
5375 recipients, not senders.
5377 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5378 the ratelimit ACL was added.
5380 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5382 PH/46 Added extra information about PostgreSQL errors to the error string.
5384 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5385 avoiding the need to install special zones in a real server. This is
5386 backwards compatible; if it can't find the fake resolver, it drops back.
5387 Thus, both old and new test suites can be run.
5389 TF/02 Added util/ratelimit.pl
5391 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5392 clock is set back in time.
5394 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5395 Candler <B.Candler@pobox.com>.
5397 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5398 OK for addresses that are the result of redirections.
5400 PH/48 A number of further additions for the benefit of the new test suite,
5401 including a fake gethostbyname() that interfaces to the fake DNS resolver
5404 TF/06 The fix for widen_domains has also been applied to qualify_single and
5405 search_parents which are the other dnslookup options that can cause
5408 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5411 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5413 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5414 service" if it encountered an MX record with an empty target hostname.
5415 The message is now "an MX or SRV record indicated no SMTP service".
5417 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5418 if the DNS of the sending site is misconfigured. This is quite a
5419 common situation. This change restores the behaviour of treating a
5420 helo verification defer as a failure.
5422 PH/51 If self=fail was set on a router, the bounce message did not include the
5423 actual error message.
5429 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5431 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5432 command itself ended in error, the underlying message about the transport
5433 filter timeout was being overwritten with the pipe command error. Now the
5434 underlying error message should be appended to the second error message.
5436 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5438 PH/02 Reduce the amount of output that "make" produces by default. Full output
5439 can still be requested.
5441 PH/03 The warning log line about a condition test deferring for a "warn" verb
5442 was being output only once per connection, rather than after each
5443 occurrence (because it was using the same function as for successful
5444 "warn" verbs). This seems wrong, so I have changed it.
5446 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5447 it should not have, which might have caused a crash in the right
5448 circumstances, but probably never did.
5450 PH/04 Installed a modified version of Tony Finch's patch to make submission
5451 mode fix the return path as well as the Sender: header line, and to
5452 add a /name= option so that you can make the user's friendly name appear
5455 TF/03 Added the control = fakedefer ACL modifier.
5457 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5458 Mark Lowes for thorough testing.
5460 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5462 TK/03 Merged latest SRS patch from Miles Wilton.
5464 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5465 with the definition in sysexits.h (which is #included earlier).
5466 Fortunately, Exim does not actually use EX_OK. The code used to try to
5467 preserve the sysexits.h value, by assuming that macro definitions were
5468 scanned for macro replacements. I have been disabused of this notion,
5469 so now the code just undefines EX_OK before #including unistd.h.
5471 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5472 in the smtp transport. When a block could not be written in a single
5473 write() function, the timeout was being re-applied to each part-write.
5474 This seems wrong - if the receiver was accepting one byte at a time it
5475 would take for ever. The timeout is now adjusted when this happens. It
5476 doesn't have to be particularly precise.
5478 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5479 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5481 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5482 without a domain if the check to <postmaster@domain> fails.
5484 SC/01 Eximstats: added -xls and the ability to specify output files
5485 (patch written by Frank Heydlauf).
5487 SC/02 Eximstats: use FileHandles for outputting results.
5489 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5491 SC/04 Eximstats: fixed display of large numbers with -nvr option
5493 SC/05 Eximstats: fixed merging of reports with empty tables.
5495 SC/06 Eximstats: added the -include_original_destination flag
5497 SC/07 Eximstats: removed tabs and trailing whitespace.
5499 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5501 TK/06 MBOX spool code: Add real "From " MBOX separator line
5502 so the .eml file is really in mbox format (even though
5503 most programs do not really care). Patch from Alex Miller.
5505 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5506 The latter is generated from $received_to and is only set if the
5507 message has one envelope recipient. SA can use these headers,
5508 obviously out-of-the-box. Patch from Alex Miller.
5510 PH/08 The ${def test on a variable was returning false if the variable's
5511 value was "0", contrary to what the specification has always said!
5512 The result should be true unless the variable is empty.
5514 PH/09 The syntax error of a character other than { following "${if
5515 def:variable_name" (after optional whitespace) was not being diagnosed.
5516 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5517 accidental colon was present, for example, could give incorrect results.
5519 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5520 result is used (not including appendfile, where other changes are about
5523 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5524 This involved changing a lot of size variables from int to off_t. It
5525 should work with maildirs and everything.
5527 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5528 spamd dying while we are connected to it.
5530 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5533 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5534 function for BDB 4.3.
5536 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5538 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5539 the sources for PCRE have been re-arranged and more files are now
5542 PH/15 The code I had for printing potentially long long variables in PH/11
5543 above was not the best (it lost precision). The length of off_t variables
5544 is now inspected at build time, and an appropriate printing format (%ld
5545 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5546 to be "long long int" or "long int". This is needed for the internal
5547 formatting function string_vformat().
5549 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5550 the configuration file to be ":syslog", then the script "guesses" where
5551 the logs files are, rather than using the compiled in default. In our
5552 case the guess is not the same as the compiled default, so the script
5553 suddenly stopped working when I started to use syslog. The patch checks
5554 to see if log_file_path is "". If so, it attempts to read it from exim
5555 with no configuration file to get the compiled in version, before it
5556 falls back to the previous guessing code."
5558 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5559 implementing BATV in an Exim configuration. See NewStuff for the gory
5562 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5563 Makefile that are specific to HP-UX.
5565 PH/18 If the "use_postmaster" option was set for a recipient callout together
5566 with the "random" option, the postmaster address was used as the MAIL
5567 FROM address for the random test, but not for the subsequent recipient
5568 test. It is now used for both.
5570 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5571 patch removes a few documentation additions to RFC 3028, because the
5572 latest draft now contains them. It adds the new en;ascii-case comparator
5573 and a new error check for 8bit text in MIME parts. Comparator and
5574 require names are now matched exactly. I enabled the subaddress
5575 extension, but it is not well tested yet (read: it works for me)."
5577 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5578 rework some of the code of TK/09 above to avoid the hardwired use of
5579 "%lld" and "long long". Replaced the call to snprintf() with a call to
5582 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5583 records point to non-existent hosts", "retry timeout exceeded", and
5584 "retry time not reached for any host after a long failure period".
5586 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5587 experimental DomainKeys support:
5589 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5590 (2) On an error such as an illegally used "control", the wrong name for
5591 the control was given.
5593 These problems did NOT occur unless DomainKeys support was compiled.
5595 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5597 PH/24 Added ${if match_ip condition.
5599 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5600 databases so that it will be absolutely obvious if a crash occurs in the
5601 DB library. This is a regular occurrence (often caused by mis-matched
5604 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5605 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5606 picked up on a user's system that detects such things. There doesn't seem
5607 to be a gcc warning option for this - only an attribute that has to be
5608 put on the function's prototype. It seems that in Fedora Core 4 they have
5609 set this on a number of new functions. No doubt there will be more in due
5612 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5613 specify a transport. However, if an address that was verified by such a
5614 router was the subject of a callout, Exim crashed because it tried to
5615 read the rcpt_include_affixes from the non-existent transport. Now it
5616 just assumes that the setting of that option is false. This bug was
5617 introduced by 4.51/PH/31.
5619 PH/28 Changed -d+all to exclude +memory, because that information is very
5620 rarely of interest, but it makes the output a lot bigger. People tend to
5621 do -d+all out of habit.
5623 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5624 code in os-type was giving problems when libc.so lives in lib64, like on
5627 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5628 aren't the modern standard, and it seems that some systems' include files
5629 don't always have them. Exim was already checking for some of the newer
5630 ones like T_AAAA, and defining it itself. I've added checks for all the
5631 record types that Exim uses.
5633 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5634 not automatically generating a new one, as it is supposed to. This
5635 prevented TLS from working. If the file did exist, but contained invalid
5636 data, a new version was generated, as expected. It was only the case of a
5637 non-existent file that was broken.
5639 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5640 with a change in libdomainkeys > 0.64.
5642 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5643 from DNS. If the selector record carries the flag, it now has
5644 precedence over the domain-wide flag.
5646 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5648 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5649 the use of a port name that isn't defined in /etc/services) occurred, the
5650 message was deferred as in a normal delivery, and thus remained on the
5651 spool, instead of being failed because of the mua_wrapper setting. This
5652 is now fixed, and I tidied up some of the mua_wrapper messages at the
5655 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5656 the messages in a hash of arrays rather than using individual hashes.
5657 This is a bit cleaner and results in dramatic memory savings, albeit
5658 at a slight CPU cost.
5660 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5661 as requested by Marc Sherman.
5663 SC/10 Eximstats: added histograms for user specified patterns as requested
5666 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5668 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5669 fopen() in the content-scanning modules that did not already have it.
5675 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5676 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5678 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5680 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5682 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5683 to test Sieve filters that use "vacation".
5685 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5686 that changes the way the GnuTLS parameters are stored in the cache file.
5687 The new format can be generated externally. For backward compatibility,
5688 if the data in the cache doesn't make sense, Exim assumes it has read an
5689 old-format file, and it generates new data and writes a new file. This
5690 means that you can't go back to an older release without removing the
5693 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5694 work if there are any delivery delays because "one_time" forces the
5695 parent to be marked "delivered", so its unseen clone is never tried
5696 again. For this reason, Exim now forbids the simultaneous setting of
5699 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5700 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5701 entry if you want to know the details. The fix, however, neglected to
5702 consider the case when local delivery batching is involved. The test for
5703 "previously delivered" was not happening when checking to see if an
5704 address could be batched with a previous (undelivered) one; under
5705 certain circumstances this could lead to multiple deliveries to the same
5708 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5709 in its include files, and this causes problems building Exim.
5711 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5712 header_syntax) but Exim was just ignoring anything given after a slash.
5713 In particular, this caused confusion with an attempt to use "verify =
5714 reverse_host_lookup/defer_ok". An error is now given when options are
5715 supplied for verify items that do not have them. (Maybe reverse_host_
5716 lookup should have a defer_ok option, but that's a different point.)
5718 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5719 defined by RFC 821) to 2048, because there were problems with some AUTH
5720 commands, and RFC 1869 says the size should be increased for extended
5721 SMTP commands that take arguments.
5723 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5726 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5727 "unknown" error; now it says that the functionality isn't in the binary.
5729 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5730 an address' error message when a string expansion fails (syntax or
5731 whatever). Otherwise the password may appear in the log. Following change
5732 PH/42 below, there is no longer a chance of it appearing in a bounce
5735 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5737 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5738 of its addresses was ever tried. (Bugzilla bug #2.)
5740 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5741 the result incorrectly in the debug output. (It correctly added a newline
5742 to what was transported.)
5744 TF/01 Added $received_time.
5746 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5747 commented out examples of how to interface to a virus scanner and to
5748 SpamAssassin. Also added commented examples of av_scanner and
5749 spamd_address settings.
5751 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5752 and controls are allowed in which ACLs. There were a couple of minor
5753 errors. Some of the entries in the conditions table (which is a table of
5754 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5755 negation of where the condition IS allowed.
5757 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5759 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5760 header file does not have a version number, so I've had to invent a new
5761 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5762 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5763 radiusclient), but it was contributed by a Radius user.
5765 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5766 files or directories, for OpenSSL.
5768 PH/19 When an Exim process that is running as root has to create an Exim log
5769 file, it does so in a subprocess that runs as exim:exim so as to get the
5770 ownership right at creation (otherwise, other Exim processes might see
5771 the file with the wrong ownership). There was no test for failure of this
5772 fork() call, which would lead to the process getting stuck as it waited
5773 for a non-existent subprocess. Forks do occasionally fail when resources
5774 run out. I reviewed all the other calls to fork(); they all seem to check
5777 PH/20 When checking for unexpected SMTP input at connect time (before writing
5778 the banner), Exim was not dealing correctly with a non-positive return
5779 from the read() function. If the client had disconnected by this time,
5780 the result was a log entry for a synchronization error with an empty
5781 string after "input=" when read() returned zero. If read() returned -1
5782 (an event I could not check), uninitialized data bytes were printed.
5783 There were reports of junk text (parts of files, etc) appearing after
5786 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5788 PH/22 Added support for macro redefinition, and (re)definition in between
5789 driver and ACL definitions.
5791 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5792 forgetting to use the resulting value; it was using the unexpanded value.
5794 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5795 hadn't been configured. The fix is from Juergen Kreileder, who
5796 understands it better than I do:
5798 "Here's what I see happening with three configured cyrus_sasl
5799 authenticators configured (plain, login, cram-md5):
5801 On startup auth_cyrus_sasl_init() gets called for each of these.
5802 This means three calls to sasl_listmech() without a specified mech_list.
5803 => SASL tests which mechs of all available mechs actually work
5804 => three warnings about OTP not working
5805 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5807 With the patch, sasl_listmech() also gets called three times. But now
5808 SASL's mech_list option is set to the server_mech specified in the the
5809 authenticator. Or in other words, the answer from sasl_listmech()
5810 gets limited to just the mech you're testing for (which is different
5812 => the return list contains just 'plain' or 'login', 'cram-md5' or
5813 nothing depending on the value of ob->server_mech.
5815 I've just tested the patch: Authentication still works fine,
5816 unavailable mechs specified in the exim configuration are still
5817 caught, and the auth.log warnings about OTP are gone."
5819 PH/25 When debugging is enabled, the contents of the command line are added
5820 to the debugging output, even when log_selector=+arguments is not
5823 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5824 answer is "GNU", and only if the return is "GNU/something" is the answer
5827 PH/27 $acl_verify_message is now set immediately after the failure of a
5828 verification in an ACL, and so is available in subsequent modifiers. In
5829 particular, the message can be preserved by coding like this:
5831 warn !verify = sender
5832 set acl_m0 = $acl_verify_message
5834 Previously, $acl_verify_message was set only while expanding "message"
5835 and "log_message" when a very denied access.
5837 PH/28 Modified OS/os.c-Linux with
5839 -#ifndef OS_LOAD_AVERAGE
5840 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5842 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5843 nomenclature these days.)
5845 PH/29 Installed patch from the Sieve maintainer that adds the options
5846 sieve_useraddress and sieve_subaddress to the redirect router.
5848 PH/30 In these circumstances:
5849 . Two addresses routed to the same list of hosts;
5850 . First host does not offer TLS;
5851 . First host accepts first address;
5852 . First host gives temporary error to second address;
5853 . Second host offers TLS and a TLS session is established;
5854 . Second host accepts second address.
5855 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5856 and peerdn, if requested) that were in fact used only for the second
5859 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5860 attention to any local part prefix or suffix that was matched by the
5861 router that accepted the address. It now behaves in the same way as it
5862 does for delivery: the affixes are removed from the local part unless
5863 rcpt_include_affixes is set on the transport.
5865 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5866 timeout during the DATA phase of an incoming message.
5868 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5869 applied a patch sanctioned by the Sieve maintainer.
5871 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5872 the uid or gid is negative. A case of a negative gid caused this to be
5873 noticed. The fix allows for either to be negative.
5875 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5876 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5879 PH/36 The change PH/12 above was broken. Fixed it.
5881 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5882 the grounds that routing the same address twice would always produce the
5883 same answer. This might have been true once, but it is certainly no
5884 longer true now. Routing a child address may depend on the previous
5885 routing that produced that child. Some complicated redirection strategies
5886 went wrong when messages had multiple recipients, and made Exim's
5887 behaviour dependent on the order in which the addresses were given.
5889 I have moved the duplicate checking until after the routing is complete.
5890 Exim scans the addresses that are assigned to local and remote
5891 transports, and removes any duplicates. This means that more work will be
5892 done, as duplicates will always all be routed, but duplicates are
5893 presumably rare, so I don't expect this is of any significance.
5895 For deliveries to pipes, files, and autoreplies, the duplicate checking
5896 still happens during the routing process, since they are not going to be
5899 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5900 It corrects a timeout issue with spamd. This is Ian's comment: "The
5901 background is that sometimes spamd either never reads data from a
5902 connection it has accepted, or it never writes response data. The exiscan
5903 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5904 blindly assumes that writes won't block so it may never time out."
5906 PH/39 Allow G after quota size as well as K and M.
5908 PH/40 The value set for $authenticated_id in an authenticator may not contain
5909 binary zeroes or newlines because the value is written to log lines and
5910 to spool files. There was no check on this. Now the value is run through
5911 the string_printing() function so that such characters are converted to
5912 printable escape sequences.
5914 PH/41 $message_linecount is a new variable that contains the total number of
5915 lines in the message. Compare $body_linecount, which is the count for the
5918 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5919 bounce and delay warning messages, except in certain special cases, which
5922 (a) An SMTP error message from a remote host;
5923 (b) A message specified in a :fail: redirection;
5924 (c) A message specified in a "fail" command in a system filter;
5925 (d) A message specified in a FAIL return from the queryprogram router;
5926 (e) A message specified by the cannot_route_message router option.
5928 In these cases only, Exim does include the error details in bounce and
5929 warning messages. There are also a few cases where bland messages such
5930 as "unrouteable address" or "local delivery error" are given.
5932 PH/43 $value is now also set for the "else" part of a ${run expansion.
5934 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5935 being worked on, but at least Exim now implements the latest version to
5938 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5939 process to complete was treated as a delivery failure, a timeout while
5940 writing the message to the pipe was logged, but erroneously treated as a
5941 successful delivery. Such timeouts include transport filter timeouts. For
5942 consistency with the overall process timeout, these timeouts are now
5943 treated as errors, giving rise to delivery failures by default. However,
5944 there is now a new Boolean option for the pipe transport called
5945 timeout_defer, which, if set TRUE, converts the failures into defers for
5946 both kinds of timeout. A transport filter timeout is now identified in
5949 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5950 systems where "make" and "gmake" are different, calling "gmake" at top
5951 level broke things. I've arranged for the value of $(MAKE) to be passed
5952 from the Makefile to this script so that it can call the same version of
5956 A note about Exim versions 4.44 and 4.50
5957 ----------------------------------------
5959 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5960 changes of various kinds. As a consequence, a big documentation update was
5961 needed. This delayed the release for rather longer than seemed good, especially
5962 in the light of a couple of (minor) security issues. Therefore, the changes
5963 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5964 release. So 4.44 and 4.50 are in effect two different branches that both start
5967 I have left the 4.50 change log unchanged; it contains all the changes since
5968 4.43. The change log for 4.44 is below; many of its items are identical to
5969 those for 4.50. This seems to be the most sensible way to preserve the
5970 historical information.
5976 1. Minor wording change to the doc/README.SIEVE file.
5978 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5979 computation of the current number of files was incorrect.
5981 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5982 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5985 4. Give more explanation in the error message when the command for a transport
5986 filter fails to execute.
5988 5. There are several places where Exim runs a non-Exim command in a
5989 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5990 was being done only for the command run by the queryprogram router. It is
5991 now done for all such subprocesses. The other cases are: ${run, transport
5992 filters, and the commands run by the lmtp and pipe transports.
5994 6. Added CONFIGURE_GROUP build-time option.
5996 7. Some older OS have a limit of 256 on the maximum number of file
5997 descriptors. Exim was using setrlimit() to set 1000 as a large value
5998 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5999 systems. I've change it so that if it can't get 1000, it tries for 256.
6001 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6002 was an oversight, and furthermore, ever since the addition of extra
6003 controls (e.g. 4.43/32), the checks on when to allow different forms of
6004 "control" were broken. There should now be diagnostics for all cases when a
6005 control that does not make sense is encountered.
6007 9. Added the /retain_sender option to "control=submission".
6009 10. $recipients is now available in the predata ACL (oversight).
6011 11. Tidy the search cache before the fork to do a delivery from a message
6012 received from the command line. Otherwise the child will trigger a lookup
6013 failure and thereby defer the delivery if it tries to use (for example) a
6014 cached ldap connection that the parent has called unbind on.
6016 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6017 of $address_data from the recipient verification was clobbered by the
6018 sender verification.
6020 13. The value of address_data from a sender verification is now available in
6021 $sender_address_data in subsequent conditions in the ACL statement.
6023 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6025 15. Added a new option "connect=<time>" to callout options, to set a different
6028 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6029 was its contents. (It was OK if the option was not defined at all.)
6031 17. A "Completed" log line is now written for messages that are removed from
6032 the spool by the -Mrm option.
6034 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6035 information about exactly what failed.
6037 19. Added -dd to debug only the daemon process.
6039 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6040 handles timeouts, both on the server side and network timeouts. Renamed the
6041 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6043 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6044 It is now set to "smtps".
6046 22. $host_address is now set to the target address during the checking of
6047 ignore_target_hosts.
6049 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6050 being passed; this would have caused $sender_host_name to have been used if
6051 matching the list had actually called for a host name (not very likely,
6052 since this list is usually IP addresses). A host name is now passed as
6055 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6056 code that specifies a non-restarting handler (typically sigaction() in
6057 modern systems) in an attempt to fix a rare and obscure crash bug.
6059 25. Narrowed the window for a race in the daemon that could cause it to ignore
6060 SIGCHLD signals. This is not a major problem, because they are used only to
6061 wake it up if nothing else does.
6063 26. A malformed maildirsize file could cause Exim to calculate negative values
6064 for the mailbox size or file count. Odd effects could occur as a result.
6065 The maildirsize information is now recalculated if the size or filecount
6068 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6069 support for a long time. Removed HAVE_SYS_VFS_H.
6071 28. Installed the latest version of exipick from John Jetmore.
6073 29. In an address list, if the pattern was not a regular expression, an empty
6074 subject address (from a bounce message) matched only if the pattern was an
6075 empty string. Non-empty patterns were not even tested. This was the wrong
6076 because it is perfectly reasonable to use an empty address as part of a
6077 database query. An empty address is now tested by patterns that are
6078 lookups. However, all the other forms of pattern expect the subject to
6079 contain a local part and a domain, and therefore, for them, an empty
6080 address still always fails if the pattern is not itself empty.
6082 30. Exim went into a mad DNS loop when attempting to do a callout where the
6083 host was specified on an smtp transport, and looking it up yielded more
6084 than one IP address.
6086 31. Re-factored the code for checking spool and log partition space into a
6087 function that finds that data and another that does the check. The former
6088 is then used to implement four new variables: $spool_space, $log_space,
6089 $spool_inodes, and $log_inodes.
6091 32. The RFC2047 encoding function was originally intended for short strings
6092 such as real names; it was not keeping to the 75-character limit for
6093 encoded words that the RFC imposes. It now respects the limit, and
6094 generates multiple encoded words if necessary. To be on the safe side, I
6095 have increased the buffer size for the ${rfc2047: expansion operator from
6098 33. It is now permitted to omit both strings after an "if" condition; if the
6099 condition is true, the result is "true". As before, when the second string
6100 is omitted, a false condition yields an empty string. This makes it less
6101 cumbersome to write custom ACL and router conditions.
6103 34. Failure to deliver a bounce message always caused it to be frozen, even if
6104 there was an errors_to setting on the router. The errors_to setting is now
6107 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6108 canonical form (fully expanded) before being placed in
6109 $sender_host_address.
6111 36. The table in the code that translates DNS record types into text (T_A to
6112 "A" for instance) was missing entries for NS and CNAME. It is just possible
6113 that this could have caused confusion if both these types were looked up
6114 for the same domain, because the text type is used as part of Exim's
6115 per-process caching. But the chance of anyone hitting this buglet seems
6118 37. The dnsdb lookup has been extended in a number of ways.
6120 (1) There is a new type, "zns", which walks up the domain tree until it
6121 finds some nameserver records. It should be used with care.
6123 (2) There is a new type, "mxh", which is like "mx" except that it returns
6124 just the host names, not the priorities.
6126 (3) It is now possible to give a list of domains (or IP addresses) to be
6127 looked up. The behaviour when one of the lookups defers can be
6128 controlled by a keyword.
6130 (4) It is now possible to specify the separator character for use when
6131 multiple records are returned.
6133 38. The dnslists ACL condition has been extended: it is now possible to supply
6134 a list of IP addresses and/or domains to be looked up in a particular DNS
6137 39. Added log_selector=+queue_time_overall.
6139 40. When running the queue in the test harness, wait just a tad after forking a
6140 delivery process, to get repeatability of debugging output.
6142 41. Include certificate and key file names in error message when GnuTLS fails
6143 to set them up, because the GnuTLS error message doesn't include the name
6144 of the failing file when there is a problem reading it.
6146 42. Allow both -bf and -bF in the same test run.
6148 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6150 44. The "Exiscan patch" is now merged into the mainline Exim source.
6152 45. Sometimes the final signoff response after QUIT could fail to get
6153 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6154 < 0 before doing a fflush(). This bug looks as though it goes back to the
6155 introduction of TLS in release 3.20, but "sometimes" must have been rare
6156 because the tests only now provoked it.
6158 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6159 (this can affect the format of dates).
6161 47. exim_tidydb, when checking for the continued existence of a message for
6162 which it has found a message-specific retry record, was not finding
6163 messages that were in split spool directories. Consequently, it was
6164 deleting retry records that should have stayed in existence.
6166 48. Steve fixed some bugs in eximstats.
6168 49. The SPA authentication driver was not abandoning authentication and moving
6169 on to the next authenticator when an expansion was forced to fail,
6170 contradicting the general specification for all authenticators. Instead it
6171 was generating a temporary error. It now behaves as specified.
6173 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6174 (the order specifies the preference for clients). The order is now AES256,
6175 AES128, 3DES, ARCFOUR128.
6177 51. Small patch to Sieve code - explicitly set From: when generating an
6180 52. Exim crashed if a remote delivery caused a very long error message to be
6181 recorded - for instance if somebody sent an entire SpamAssassin report back
6182 as a large number of 550 error lines. This bug was coincidentally fixed by
6183 increasing the size of one of Exim's internal buffers (big_buffer) that
6184 happened as part of the Exiscan merge. However, to be on the safe side, I
6185 have made the code more robust (and fixed the comments that describe what
6188 53. Now that there can be additional text after "Completed" in log lines (if
6189 the queue_time_overall log selector is set), a one-byte patch to exigrep
6190 was needed to allow it to recognize "Completed" as not the last thing in
6193 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6194 patch that reportedly fixes this has been added. I am not expert enough to
6195 create a test for it. This is what the patch creator wrote:
6197 "I found a little strange behaviour of ldap code when working with
6198 Windows 2003 AD Domain, where users was placed in more than one
6199 Organization Units. When I tried to give exim partial DN, the exit code
6200 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6201 But simultaneously result of request was absolutely normal ldap result,
6202 so I produce this patch..."
6204 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6205 REFERENCE, so I have modified the code to exclude the patch when that macro
6208 55. Some experimental protocols are using DNS PTR records for new purposes. The
6209 keys for these records are domain names, not reversed IP addresses. The
6210 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6211 leaves it alone. Component reversal etc. now happens only for IP addresses.
6214 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6216 57. Double the size of the debug message buffer (to 2048) so that more of very
6217 long debug lines gets shown.
6219 58. The exicyclog utility now does better if the number of log files to keep
6220 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6222 59. Two changes related to the smtp_active_hostname option:
6224 (1) $smtp_active_hostname is now available as a variable.
6225 (2) The default for smtp_banner uses $smtp_active_hostname instead
6226 of $primary_hostname.
6228 60. The host_aton() function is supposed to be passed a string that is known
6229 to be a valid IP address. However, in the case of IPv6 addresses, it was
6230 not checking this. This is a hostage to fortune. Exim now panics and dies
6231 if the condition is not met. A case was found where this could be provoked
6232 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6233 components; fortuitously, this particular loophole had already been fixed
6234 by change 4.50/55 above.
6236 If there are any other similar loopholes, the new check in host_aton()
6237 itself should stop them being exploited. The report I received stated that
6238 data on the command line could provoke the exploit when Exim was running as
6239 exim, but did not say which command line option was involved. All I could
6240 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6241 running as the user.
6244 61. There was a buffer overflow vulnerability in the SPA authentication code
6245 (which came originally from the Samba project). I have added a test to the
6246 spa_base64_to_bits() function which I hope fixes it.
6249 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6250 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6252 63. The daemon start-up calls getloadavg() while still root for those OS that
6253 need the first call to be done as root, but it missed one case: when
6254 deliver_queue_load_max is set with deliver_drop_privilege. This is
6255 necessary for the benefit of the queue runner, because there is no re-exec
6256 when deliver_drop_privilege is set.
6258 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6259 This has been fixed.
6261 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6262 was in use, was not putting the data itself into the right store pool;
6263 consequently, it could be overwritten for a subsequent message in the same
6264 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6267 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6269 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6270 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6271 the function were treating the return as a boolean value, which happened to
6272 work because 0=false and not-0=true, but is not correct code.
6274 68. The host_aton() function was not handling scoped IPv6 addresses (those
6275 with, for example, "%eth0" on the end) correctly.
6277 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6278 negated items (that is, ~something) in unsigned ints. Some compilers
6279 apparently mutter when there is no cast.
6281 70. If an address verification called from an ACL failed, and did not produce a
6282 user-specific message (i.e. there was only a "system" message), nothing was
6283 put in $acl_verify_message. In this situation, it now puts the system
6286 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6287 session; change 4.31/43 added the unwanted input to the log line - except
6288 that it did not do this in the start of session case. It now does.
6290 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6291 This is wrong and can cause the other end to generate a synchronization
6292 error if it is another Exim or anything else that does the synchronization
6293 check. A QUIT command is no longer sent after a timeout.
6295 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6296 during host lookups.
6298 74. The defer_ok option of callout verification was not working if it was used
6299 when verifying addresses in header lines, that is, for this case:
6301 verify = header_sender/callout=defer_ok
6303 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6304 those file descriptors could be used for SMTP connections. If anything
6305 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6306 could be sent to the SMTP client, causing chaos. The daemon now opens
6307 stdin, stdout, and stderr to /dev/null when it puts itself into the
6310 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6311 log by default. The user can override this with suitable Perl magic.
6313 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6314 the log message when discard triggers, was not working for the DATA ACL or
6315 for the non-SMTP ACL.
6317 78. Error message wording change in sieve.c.
6319 79. If smtp_accept_max_per_host was set, the number of connections could be
6320 restricted to fewer than expected, because the daemon was trying to set up
6321 a new connection before checking whether the processes handling previous
6322 connections had finished. The check for completed processes is now done
6323 earlier. On busy systems, this bug wouldn't be noticed because something
6324 else would have woken the daemon, and it would have reaped the completed
6327 80. If a message was submitted locally by a user whose login name contained one
6328 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6329 It caused a spool format error. I have fixed the spool reading code. A
6330 related problem was that the "from" clause in the Received: line became
6331 illegal because of the space(s). It is now covered by ${quote_local_part.
6333 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6336 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6338 83. Patch from Sieve maintainer to fix unterminated string problem in
6339 "vacation" handling.
6341 84. Some minor changes to the Linux configuration files to help with other
6342 OS variants using glibc.
6344 85. One more patch for Sieve to update vacation handling to latest spec.
6347 ----------------------------------------------------
6348 See the note above about the 4.44 and 4.50 releases.
6349 ----------------------------------------------------
6355 1. Change 4.43/35 introduced a bug that caused file counts to be
6356 incorrectly computed when quota_filecount was set in an appendfile
6359 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6360 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6363 3. Give more explanation in the error message when the command for a transport
6364 filter fails to execute.
6366 4. There are several places where Exim runs a non-Exim command in a
6367 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6368 was being done only for the command run by the queryprogram router. It is
6369 now done for all such subprocesses. The other cases are: ${run, transport
6370 filters, and the commands run by the lmtp and pipe transports.
6372 5. Some older OS have a limit of 256 on the maximum number of file
6373 descriptors. Exim was using setrlimit() to set 1000 as a large value
6374 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6375 systems. I've change it so that if it can't get 1000, it tries for 256.
6377 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6378 was an oversight, and furthermore, ever since the addition of extra
6379 controls (e.g. 4.43/32), the checks on when to allow different forms of
6380 "control" were broken. There should now be diagnostics for all cases when a
6381 control that does not make sense is encountered.
6383 7. $recipients is now available in the predata ACL (oversight).
6385 8. Tidy the search cache before the fork to do a delivery from a message
6386 received from the command line. Otherwise the child will trigger a lookup
6387 failure and thereby defer the delivery if it tries to use (for example) a
6388 cached ldap connection that the parent has called unbind on.
6390 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6391 of $address_data from the recipient verification was clobbered by the
6392 sender verification.
6394 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6395 was its contents. (It was OK if the option was not defined at all.)
6397 11. A "Completed" log line is now written for messages that are removed from
6398 the spool by the -Mrm option.
6400 12. $host_address is now set to the target address during the checking of
6401 ignore_target_hosts.
6403 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6404 being passed; this would have caused $sender_host_name to have been used if
6405 matching the list had actually called for a host name (not very likely,
6406 since this list is usually IP addresses). A host name is now passed as
6409 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6410 code that specifies a non-restarting handler (typically sigaction() in
6411 modern systems) in an attempt to fix a rare and obscure crash bug.
6413 15. Narrowed the window for a race in the daemon that could cause it to ignore
6414 SIGCHLD signals. This is not a major problem, because they are used only to
6415 wake it up if nothing else does.
6417 16. A malformed maildirsize file could cause Exim to calculate negative values
6418 for the mailbox size or file count. Odd effects could occur as a result.
6419 The maildirsize information is now recalculated if the size or filecount
6422 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6423 support for a long time. Removed HAVE_SYS_VFS_H.
6425 18. Updated exipick to current release from John Jetmore.
6427 19. Allow an empty sender to be matched against a lookup in an address list.
6428 Previously the only cases considered were a regular expression, or an
6431 20. Exim went into a mad DNS lookup loop when doing a callout where the
6432 host was specified on the transport, if the DNS lookup yielded more than
6435 21. The RFC2047 encoding function was originally intended for short strings
6436 such as real names; it was not keeping to the 75-character limit for
6437 encoded words that the RFC imposes. It now respects the limit, and
6438 generates multiple encoded words if necessary. To be on the safe side, I
6439 have increased the buffer size for the ${rfc2047: expansion operator from
6442 22. Failure to deliver a bounce message always caused it to be frozen, even if
6443 there was an errors_to setting on the router. The errors_to setting is now
6446 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6447 canonical form (fully expanded) before being placed in
6448 $sender_host_address.
6450 24. Updated eximstats to version 1.33
6452 25. Include certificate and key file names in error message when GnuTLS fails
6453 to set them up, because the GnuTLS error message doesn't include the name
6454 of the failing file when there is a problem reading it.
6456 26. Expand error message when OpenSSL has problems setting up cert/key files.
6459 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6460 (this can affect the format of dates).
6462 28. exim_tidydb, when checking for the continued existence of a message for
6463 which it has found a message-specific retry record, was not finding
6464 messages that were in split spool directories. Consequently, it was
6465 deleting retry records that should have stayed in existence.
6467 29. eximstats updated to version 1.35
6468 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6469 1.35 - bugfix such that pie charts by volume are generated correctly
6471 30. The SPA authentication driver was not abandoning authentication and moving
6472 on to the next authenticator when an expansion was forced to fail,
6473 contradicting the general specification for all authenticators. Instead it
6474 was generating a temporary error. It now behaves as specified.
6476 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6477 (the order specifies the preference for clients). The order is now AES256,
6478 AES128, 3DES, ARCFOUR128.
6480 31. Small patch to Sieve code - explicitly set From: when generating an
6483 32. Exim crashed if a remote delivery caused a very long error message to be
6484 recorded - for instance if somebody sent an entire SpamAssassin report back
6485 as a large number of 550 error lines. This bug was coincidentally fixed by
6486 increasing the size of one of Exim's internal buffers (big_buffer) that
6487 happened as part of the Exiscan merge. However, to be on the safe side, I
6488 have made the code more robust (and fixed the comments that describe what
6491 33. Some experimental protocols are using DNS PTR records for new purposes. The
6492 keys for these records are domain names, not reversed IP addresses. The
6493 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6494 leaves it alone. Component reversal etc. now happens only for IP addresses.
6497 34. The host_aton() function is supposed to be passed a string that is known
6498 to be a valid IP address. However, in the case of IPv6 addresses, it was
6499 not checking this. This is a hostage to fortune. Exim now panics and dies
6500 if the condition is not met. A case was found where this could be provoked
6501 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6502 components; fortuitously, this particular loophole had already been fixed
6503 by change 4.50/55 or 4.44/33 above.
6505 If there are any other similar loopholes, the new check in host_aton()
6506 itself should stop them being exploited. The report I received stated that
6507 data on the command line could provoke the exploit when Exim was running as
6508 exim, but did not say which command line option was involved. All I could
6509 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6510 running as the user.
6513 35. There was a buffer overflow vulnerability in the SPA authentication code
6514 (which came originally from the Samba project). I have added a test to the
6515 spa_base64_to_bits() function which I hope fixes it.
6518 36. The daemon start-up calls getloadavg() while still root for those OS that
6519 need the first call to be done as root, but it missed one case: when
6520 deliver_queue_load_max is set with deliver_drop_privilege. This is
6521 necessary for the benefit of the queue runner, because there is no re-exec
6522 when deliver_drop_privilege is set.
6524 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6525 was in use, was not putting the data itself into the right store pool;
6526 consequently, it could be overwritten for a subsequent message in the same
6527 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6530 38. Sometimes the final signoff response after QUIT could fail to get
6531 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6532 < 0 before doing a fflush(). This bug looks as though it goes back to the
6533 introduction of TLS in release 3.20, but "sometimes" must have been rare
6534 because the tests only now provoked it.
6540 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6541 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6542 Now it returns an integer. A number of calls were still expecting a T/F
6543 return. Fortuitously, in all cases, the tests worked in OK situations,
6544 which is the norm. However, things would have gone wrong on any write
6545 failures on the smtp file descriptor. This function is used when sending
6546 messages over SMTP and also when doing verify callouts.
6548 2. When Exim is called to do synchronous delivery of a locally submitted
6549 message (the -odf or -odi options), it no longer closes stderr before doing
6552 3. Implemented the mua_wrapper option.
6554 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6556 5. Implemented the functions header_remove(), header_testname(),
6557 header_add_at_position(), and receive_remove_recipient(), and exported them
6560 6. If an ACL "warn" statement specified the addition of headers, Exim already
6561 inserted X-ACL-Warn: at the start if there was no header name. However, it
6562 was not making this test for the second and subsequent header lines if
6563 there were newlines in the string. This meant that an invalid header could
6564 be inserted if Exim was badly configured.
6566 7. Allow an ACL "warn" statement to add header lines at the start or after all
6567 the Received: headers, as well as at the end.
6569 8. Added the rcpt_4xx retry error code.
6571 9. Added postmaster_mailfrom=xxx to callout verification option.
6573 10. Added mailfrom=xxxx to the callout verification option, for verify=
6576 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6577 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6579 12. Inserted some casts to stop certain compilers warning when using pointer
6580 differences as field lengths or precisions in printf-type calls (mostly
6581 affecting debugging statements).
6583 13. Added optional readline() support for -be (dynamically loaded).
6585 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6586 same clock tick as a message's arrival, so that its received time was the
6587 same as the "first fail" time on the retry record, and that message
6588 remained on the queue past the ultimate address timeout, every queue runner
6589 would try a delivery (because it was past the ultimate address timeout) but
6590 after another failure, the ultimate address timeout, which should have then
6591 bounced the address, did not kick in. This was a "< instead of <=" error;
6592 in most cases the first failure would have been in the next clock tick
6593 after the received time, and all would be well.
6595 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6596 being recognized when the domain list was tested by the match_domain
6597 condition in an expansion string.
6599 16. Added the ${str2b64: operator.
6601 17. Exim was always calling setrlimit() to set a large limit for the number of
6602 processes, without checking whether the existing limit was already
6603 adequate. (It did check for the limit on file descriptors.) Furthermore,
6604 errors from getrlimit() and setrlimit() were being ignored. Now they are
6605 logged to the main and panic logs, but Exim does carry on, to try to do its
6606 job under whatever limits there are.
6608 18. Imported PCRE 5.0.
6610 19. Trivial typo in log message " temporarily refused connection" (the leading
6613 20. If the log selector return_path_on_delivery was set and an address was
6614 redirected to /dev/null, the delivery process crashed because it assumed
6615 that a return path would always be set for a "successful" delivery. In this
6616 case, the whole delivery is bypassed as an optimization, and therefore no
6619 21. Internal re-arrangement: the function for sending a challenge and reading
6620 a response while authentication was assuming a zero-terminated challenge
6621 string. It's now changed to take a pointer and a length, to allow for
6622 binary data in such strings.
6624 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6626 23. Exim was not respecting finduser_retries when seeking the login of the
6627 uid under which it was called; it was always trying 10 times. (The default
6628 setting of finduser_retries is zero.) Also, it was sleeping after the final
6629 failure, which is pointless.
6631 24. Implemented tls_on_connect_ports.
6633 25. Implemented acl_smtp_predata.
6635 26. If the domain in control=submission is set empty, Exim assumes that the
6636 authenticated id is a complete email address when it generates From: or
6637 Sender: header lines.
6639 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6640 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6641 chown and chgrp in /bin and hostname in /usr/bin.
6643 28. Exim was keeping the "process log" file open after each use, just as it
6644 does for the main log. This opens the possibility of it remaining open for
6645 long periods when the USR1 signal hits a daemon. Occasional processlog
6646 errors were reported, that could have been caused by this. Anyway, it seems
6647 much more sensible not to leave this file open at all, so that is what now
6650 29. The long-running daemon process does not normally write to the log once it
6651 has entered its main loop, and it closes the log before doing so. This is
6652 so that log files can straightforwardly be renamed and moved. However,
6653 there are a couple of unusual error situations where the daemon does write
6654 log entries, and I had neglected to close the log afterwards.
6656 30. The text of an SMTP error response that was received during a remote
6657 delivery was being truncated at 512 bytes. This is too short for some of
6658 the long messages that one sometimes sees. I've increased the limit to
6661 31. It is now possible to make retry rules that apply only when a message has a
6662 specific sender, in particular, an empty sender.
6664 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6665 it possible to be selective about when SMTP synchronization is enforced.
6667 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6669 32. Implemented hosts_connection_nolog.
6671 33. Added an ACL for QUIT.
6673 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6676 35. Added mailbox_size and mailbox_filecount to appendfile.
6678 36. Added control = no_multiline_responses to ACLs.
6680 37. There was a bug in the logic of the code that waits for the clock to tick
6681 in the case where the clock went backwards by a substantial amount such
6682 that the microsecond fraction of "now" was more than the microsecond
6683 fraction of "then" (but the whole seconds number was less).
6685 38. Added support for the libradius Radius client library this is found on
6686 FreeBSD (previously only the radiusclient library was supported).
6692 1. When certain lookups returned multiple values in the form name=value, the
6693 quoting of the values was not always being done properly. Specifically:
6694 (a) If the value started with a double quote, but contained no whitespace,
6696 (b) If the value contained whitespace other than a space character (i.e.
6697 tabs or newlines or carriage returns) it was not quoted.
6698 This fix has been applied to the mysql and pgsql lookups by writing a
6699 separate quoting function and calling it from the lookup code. The fix
6700 should probably also be applied to nisplus, ibase and oracle lookups, but
6701 since I cannot test any of those, I have not disturbed their existing code.
6703 2. A hit in the callout cache for a specific address caused a log line with no
6704 reason for rejecting RCPT. Now it says "Previous (cached) callout
6705 verification failure".
6707 3. There was an off-by-one bug in the queryprogram router. An over-long
6708 return line was truncated at 256 instead of 255 characters, thereby
6709 overflowing its buffer with the terminating zero. As well as fixing this, I
6710 have increased the buffer size to 1024 (and made a note to document this).
6712 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6713 when Exim is waiting for an SMTP response from a remote server, Exim
6714 restarts its select() call on the socket, thereby resetting its timeout.
6715 This is not a problem when such interrupts are rare. Somebody set up a cron
6716 job to run exiwhat every 2 minutes, which is less than the normal select()
6717 timeout (5 or 10 minutes). This meant that the select() timeout never
6718 kicked in because it was always reset. I have fixed this by comparing the
6719 time when an interrupt arrives with the time at the start of the first call
6720 to select(). If more time than the timeout has elapsed, the interrupt is
6721 treated as a timeout.
6723 5. Some internal re-factoring in preparation for the addition of Sieve
6724 extensions (by MH). In particular, the "personal" test is moved to a
6725 separate function, and given an option for scanning Cc: and Bcc: (which is
6726 not set for Exim filters).
6728 6. When Exim created an email address using the login of the caller as the
6729 local part (e.g. when creating a From: or Sender: header line), it was not
6730 quoting the local part when it contained special characters such as @.
6732 7. Installed new OpenBSD configuration files.
6734 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6735 try to make them clearer.
6737 9. Callout options, other than the timeout value, were being ignored when
6738 verifying sender addresses in header lines. For example, when using
6740 verify = header_sender/callout=no_cache
6742 the cache was (incorrectly) being used.
6744 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6745 only the Cygwin environment.
6747 11. When return_path_on_delivery was set as a log selector, if different remote
6748 addresses in the same message used different return paths and parallel
6749 remote delivery occurred, the wrong values would sometimes be logged.
6750 (Whenever a remote delivery process finished, the return path value from
6751 the most recently started remote delivery process was logged.)
6753 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6754 lines when AUTH and/or TLS are in use. This is the "received protocol"
6755 field. Exim used to use "asmtp" for authenticated SMTP, without any
6756 indication (in the protocol name) for TLS use. Now it follows the RFC and
6757 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6758 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6759 names appear in log lines as well as in Received: header lines.
6761 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6762 extensions, and comparison tests, and to fix some bugs.
6764 14. Changes to the "personal" filter test:
6766 (1) The test was buggy in that it was just doing the equivalent of
6767 "contains" tests on header lines. For example, if a user's address was
6768 anne@some.where, the "personal" test would incorrectly be true for
6770 To: susanne@some.where
6772 This test is now done by extracting each address from the header in turn,
6773 and checking the entire address. Other tests that are part of "personal"
6774 are now done using regular expressions (for example, to check local parts
6775 of addresses in From: header lines).
6777 (2) The list of non-personal local parts in From: addresses has been
6778 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6779 taken from the Sieve specification recommendations.
6781 (3) If the message contains any header line starting with "List-" it is
6782 treated as non-personal.
6784 (4) The test for "circular" in the Subject: header line has been removed
6785 because it now seems ill-conceived.
6787 15. Minor typos in src/EDITME comments corrected.
6789 16. Installed latest exipick from John Jetmore.
6791 17. If headers_add on a router specified a text string that was too long for
6792 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6793 of string_sprintf() is now avoided.
6795 18. $message_body_size was not set (it was always zero) when running the DATA
6796 ACL and the local_scan() function.
6798 19. For the "mail" command in an Exim filter, no default was being set for
6799 the once_repeat time, causing a random time value to be used if "once" was
6800 specified. (If the value happened to be <= 0, no repeat happened.) The
6801 default is now 0s, meaning "never repeat". The "vacation" command was OK
6802 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6803 (I found it when inspecting the code).
6805 20. There is now an overall timeout for performing a callout verification. It
6806 defaults to 4 times the callout timeout, which applies to individual SMTP
6807 commands during the callout. The overall timeout applies when there is more
6808 than one host that can be tried. The timeout is checked before trying the
6809 next host. This prevents very long delays if there are a large number of
6810 hosts and all are timing out (e.g. when the network connections are timing
6811 out). The value of the overall timeout can be changed by specifying an
6812 additional sub-option for "callout", called "maxwait". For example:
6814 verify = sender/callout=5s,maxwait=20s
6816 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6817 to the end before writing, but this should make it even safer).
6819 22. Exim was forgetting that it had advertised PIPELINING for the second and
6820 subsequent messages on an SMTP connection. It was also not resetting its
6821 memory on STARTTLS and an internal HELO.
6823 23. When Exim logs an SMTP synchronization error within a session, it now
6824 records whether PIPELINING has been advertised or not.
6826 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6827 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6828 rather than long int.
6830 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6832 26. Added the never_mail option to autoreply.
6838 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6839 crash if the getsockname() call failed; this can happen if a connection is
6840 closed very soon after it is established. The problem was simply in the
6841 order in which certain operations were done, causing Exim to try to write
6842 to the SMTP stream before it had set up the file descriptor. The bug has
6843 been fixed by making things happen in the correct order.
6849 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6850 before the connection was closed, thus losing the rejection response.
6852 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6853 some early Solaris releases, but causes trouble in current releases where
6854 socklen_t is defined.
6856 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6859 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6862 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6863 that releasing the top of it at the end releases what was used for sub-
6864 expansions (unless the block got too big). However, discard this block if
6865 the first thing is a variable or header, so that we can use its block when
6866 it is dynamic (useful for very large $message_headers, for example).
6868 6. Lookups now cache *every* query, not just the most recent. A new, separate
6869 store pool is used for this. It can be recovered when all lookup caches are
6870 flushed. Lookups now release memory at the end of their result strings.
6871 This has involved some general refactoring of the lookup sources.
6873 7. Some code has been added to the store_xxx() functions to reduce the amount
6874 of flapping under certain conditions.
6876 8. log_incoming_interface used to affect only the <= reception log lines. Now
6877 it causes the local interface and port to be added to several more SMTP log
6878 lines, for example "SMTP connection from", and rejection lines.
6880 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6882 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6884 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6885 could be overwritten at the end of the current message (or the start of a
6886 new message if it was set in a HELO ACL). The value is now preserved for
6887 the duration of the SMTP connection.
6889 12. If a transport had a headers_rewrite setting, and a matching header line
6890 contained an unqualified address, that address was qualified, even if it
6891 did not match any rewriting rules. The underlying bug was that the values
6892 of the flags that permit the existence of unqualified sender and recipient
6893 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6894 non-local messages, and by -bnq for local messages) were not being
6895 preserved with the message after it was received.
6897 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6898 "next input=" as part of the text comprising the host identity instead of
6899 the correct text. The code was using the same buffer for two different
6900 strings. However, depending on which order the printing function evaluated
6901 its arguments, the bug did not always show up. Under Linux, for example, my
6902 test suite worked just fine.
6904 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6905 doesn't work with some older versions of Perl. It has been changed to "my",
6906 which in any case is probably the better facility to use.
6908 15. A really picky compiler found some instances of statements for creating
6909 error messages that either had too many or two few arguments for the format
6912 16. The size of the buffer for calls to the DNS resolver has been increased
6913 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6914 for addresses that have a lot of PTR records. This alleviates a problem; it
6915 does not fully solve it.
6917 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6918 buffer now truncates the list and logs the incident, which is the same
6919 action as happens when Exim is looking up a host name and its aliases.
6920 Previously in this situation something unpredictable would happen;
6921 sometimes it was "internal error: store_reset failed".
6923 18. If a server dropped the connection unexpectedly when an Exim client was
6924 using GnuTLS and trying to read a response, the client delivery process
6925 crashed while trying to generate an error log message.
6927 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6928 string, for example:
6930 warn message = H1: something\nH2: something
6932 the text was added as a single header line from Exim's point of view
6933 though it ended up OK in the delivered message. However, searching for the
6934 second and subsequent header lines using $h_h2: did not work. This has been
6935 fixed. Similarly, if a system filter added multiple headers in this way,
6936 the routers could not see them.
6938 20. Expanded the error message when iplsearch is called with an invalid key to
6939 suggest using net-iplsearch in a host list.
6941 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6942 ACLs are no longer actually imposed (and a message to that effect is
6945 22. If a "gecos" field in a passwd entry contained escaped characters, in
6946 particular, if it contained a \" sequence, Exim got it wrong when building
6947 a From: or a Sender: header from that name. A second bug also caused
6948 incorrect handling when an unquoted " was present following a character
6949 that needed quoting.
6951 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6952 was not being matched caselessly.
6954 24. Arranged for all hyphens in the exim.8 source to be escaped with
6957 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6958 properly. Recipient callout cache records were still being keyed to include
6959 the sender, even when use_sender was set false. This led to far more
6960 callouts that were necessary. The sender is no longer included in the key
6961 when use_sender is false.
6963 26. Added "control = submission" modifier to ACLs.
6965 27. Added the ${base62d: operator to decode base 62 numbers.
6967 28. dnsdb lookups can now access SRV records.
6969 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6970 the configuration file.
6972 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6973 (-v) mode. This makes the output for a verbose queue run more intelligible.
6975 31. Added a use_postmaster feature to recipient callouts.
6977 32. Added the $body_zerocount variable, containing the number of binary zero
6978 bytes in the message body.
6980 33. The time of last modification of the "new" subdirectory is now used as the
6981 "mailbox time last read" when there is a quota error for a maildir
6984 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6986 35. Added +ignore_unknown as a special item in host lists.
6988 36. Code for decoding IPv6 addresses in host lists is now included, even if
6989 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6990 address was recognized as an IP address, but was then not correctly decoded
6991 into binary, causing unexpected and incorrect effects when compared with
6998 1. Very minor rewording of debugging text in manualroute to say "list of
6999 hosts" instead of "hostlist".
7001 2. If verify=header_syntax was set, and a header line with an unqualified
7002 address (no domain) and a large number of spaces between the end of the
7003 name and the colon was received, the reception process suffered a buffer
7004 overflow, and (when I tested it) crashed. This was caused by some obsolete
7005 code that should have been removed. The fix is to remove it!
7007 3. When running in the test harness, delay a bit after writing a bounce
7008 message to get a bit more predictability in the log output.
7010 4. Added a call to search_tidyup() just before forking a reception process. In
7011 theory, someone could use a lookup in the expansion of smtp_accept_max_
7012 per_host which, without the tidyup, could leave open a database connection.
7014 5. Added the variables $recipient_data and $sender_data which get set from a
7015 lookup success in an ACL "recipients" or "senders" condition, or a router
7016 "senders" option, similar to $domain_data and $local_part_data.
7018 6. Moved the writing of debug_print from before to after the "senders" test
7021 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7022 problems for message scanning, either using a data ACL, or using
7023 local_scan() because the Received: header was not generated till after they
7024 were called (in order to set the time as the time of reception completion).
7025 I have revised the way this works. The header is now generated after the
7026 body is received, but before the ACL or local_scan() are called. After they
7027 are run, the timestamp in the header is updated.
7033 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7034 before starting a queue runner without re-exec. This happened only when
7035 deliver_drop_privilege was set or when the Exim user was set to root. The
7036 effect of the bug was that timeouts during subsequent deliveries caused
7037 crashes instead of being properly handled. The handler is now left at its
7038 default (and expected) setting.
7040 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7041 message, again when deliver_drop_privilege is set or Exim is run as root.
7042 The bug described in (1) was not present in this case, but the tidying up
7043 of the other signals was missing. I have made the two cases consistent.
7045 3. The ignore_target_hosts setting on a manualroute router was being ignored
7046 for hosts that were looked up using the /MX notation.
7048 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7051 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7052 operated on the sender address. After changing the $sender_address to <>
7053 for the sender address verify, Exim was re-instated it as the original
7054 (before rewriting) address, but remembering that it had rewritten it, so it
7055 wasn't rewriting it again. This bug also had the effect of breaking the
7056 sender address verification caching when the sender address was rewritten.
7058 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7059 This has been changed so that if the ip literal address matches
7060 ignore_target_hosts, the router declines.
7062 7. Added expansion conditions match_domain, match_address, and match_local_
7063 part (NOT match_host).
7065 8. The placeholder for the Received: header didn't have a length field set.
7067 9. Added code to Exim itself and to exim_lock to test for a specific race
7068 condition that could lead to file corruption when using MBX delivery. The
7069 issue is with the lockfile that is created in /tmp. If this file is removed
7070 after a process has opened it but before that process has acquired a lock,
7071 there is the potential for a second process to recreate the file and also
7072 acquire a lock. This could lead to two Exim processes writing to the file
7073 at the same time. The added code performs the same test as UW imapd; it
7074 checks after acquiring the lock that its file descriptor still refers to
7075 the same named file.
7077 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7078 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7081 11. Added the smtp_active_hostname option. If used, this will typically be made
7082 to depend on the incoming interface address. Because $interface_address is
7083 not set up until the daemon has forked a reception process, error responses
7084 that can happen earlier (such as "too many connections") no longer contain
7087 12. If an expansion in a condition on a "warn" statement fails because a lookup
7088 defers, the "warn" statement is abandoned, and the next ACL statement is
7089 processed. Previously this caused the whole ACL to be aborted.
7091 13. Added the iplsearch lookup type.
7093 14. Added ident_timeout as a log selector.
7095 15. Added tls_certificate_verified as a log selector.
7097 16. Added a global option tls_require_ciphers (compare the smtp transport
7098 option of the same name). This controls incoming TLS connections.
7100 17. I finally figured out how to make tls_require_ciphers do a similar thing
7101 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7102 before starting the TLS session.
7104 18. Tabs are now shown as \t in -bP output.
7106 19. If the log selector return_path_on_delivery was set, Exim crashed when
7107 bouncing a message because it had too many Received: header lines.
7109 20. If two routers both had headers_remove settings, and the first one included
7110 a superfluous trailing colon, the final name in the first list and the
7111 first name in the second list were incorrectly joined into one item (with a
7112 colon in the middle).
7118 1. Added -C and -D options to the exinext utility, mainly to make it easier
7119 to include in the automated testing, but these could be helpful when
7120 multiple configurations are in use.
7122 2. The exinext utility was not formatting the output nicely when there was
7123 an alternate port involved in the retry record key, nor when there was a
7124 message id as well (for retries that were specific to a specific message
7125 and a specific host). It was also confused by IPv6 addresses, because of
7126 the additional colons they contain. I have fixed the IPv4 problem, and
7127 patched it up to do a reasonable job for IPv6.
7129 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7130 delivery, the log line now contains "pipelined" if PIPELINING was used.
7132 4. An SMTP transport process used to panic and die if the bind() call to set
7133 an explicit outgoing interface failed. This has been changed; it is now
7134 treated in the same way as a connect() failure.
7136 5. A reference to $sender_host_name in the part of a conditional expansion
7137 that was being skipped was still causing a DNS lookup. This no longer
7140 6. The def: expansion condition was not recognizing references to header lines
7141 that used bh_ and bheader_.
7143 7. Added the _cache feature to named lists.
7145 8. The code for checking quota_filecount in the appendfile transport was
7146 allowing one more file than it should have been.
7148 9. For compatibility with Sendmail, the command line option
7156 and sets the incoming protocol and host name (for trusted callers). The
7157 host name and its colon can be omitted when only the protocol is to be set.
7158 Note the Exim already has two private options, -pd and -ps, that refer to
7159 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7160 "s", but I don't think that's a major issue.
7162 10. A number of refactoring changes to the code, none of which should affect
7165 (a) The number of logging options was getting close to filling up the
7166 32-bit word that was used as a bit map. I have split them into two classes:
7167 those that are passed in the argument to log_write(), and those that are
7168 only ever tested independently outside of that function. These are now in
7169 separate 32-bit words, so there is plenty of room for expansion again.
7170 There is no change in the user interface or the logging behaviour.
7172 (b) When building, for example, log lines, the code previously used a
7173 macro that called string_cat() twice, in order to add two strings. This is
7174 not really sufficiently general. Furthermore, there was one instance where
7175 it was actually wrong because one of the argument was used twice, and in
7176 one call a function was used. (As it happened, calling the function twice
7177 did not affect the overall behaviour.) The macro has been replaced by a
7178 function that can join an arbitrary number of extra strings onto a growing
7181 (c) The code for expansion conditions now uses a table and a binary chop
7182 instead of a serial search (which was left over from when there were very
7183 few conditions). Also, it now recognizes conditions like "pam" even when
7184 the relevant support is not compiled in: a suitably worded error message is
7185 given if an attempt is made to use such a condition.
7187 11. Added ${time_interval:xxxxx}.
7189 12. A bug was causing one of the ddress fields not to be passed back correctly
7190 from remote delivery subprocesses. The field in question was not being
7191 subsequently used, so this caused to problems in practice.
7193 13. Added new log selectors queue_time and deliver_time.
7195 14. Might have fixed a bug in maildirsizefile handling that threw up
7196 "unexpected character" debug warnings, and recalculated the data
7197 unnecessarily. In any case, I expanded the warning message to give more
7200 15. Added the message "Restricted characters in address" to the statements in
7201 the default ACL that block characters like @ and % in local parts.
7203 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7204 Three changes have been made:
7206 (a) There was a serious bug; a negative response to MAIL caused the whole
7207 recipient domain to be cached as invalid, thereby blocking all messages
7208 to all local parts at the same domain, from all senders. This bug has
7209 been fixed. The domain is no longer cached after a negative response to
7210 MAIL if the sender used is not empty.
7212 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7215 (c) A new callout option, "use_sender" has been added for people who want
7216 the modified behaviour.
7222 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7225 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7226 indeed breaks things for older releases.
7228 3. Added additional logging to the case where there is a problem reading data
7229 from a filter that is running in a subprocess using a pipe, in order to
7230 try to track down a specific problem.
7232 4. Testing facility fudge: when running in the test harness and attempting
7233 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7234 getting "No route to host". Convert this to a timeout.
7236 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7239 6. Some OS don't have socklen_t but use size_t instead. This affects the
7240 fifth argument of getsockopt() amongst other things. This is now
7241 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7242 can be set for individual OS. I have set it for SunOS5, OSF1, and
7243 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7244 some earlier ones do not.
7246 7. Change 4.30/15 was not doing the test caselessly.
7248 8. The standard form for an IPv6 address literal was being rejected by address
7249 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7250 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7251 this, as well as the form without the "IPv6" on the front (but only when
7252 address literals are enabled, of course).
7254 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7256 10. Exim crashed if a message with an empty sender address specified by -f
7257 encountered a router with an errors_to setting. This could be provoked only
7258 by a command such as
7262 where an empty string was supplied; "<>" did not hit this bug.
7264 11. Installed PCRE release 4.5.
7266 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7267 remained set. It is now erased.
7269 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7270 times from message ids (which are base 36 rather than the normal 62).
7272 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7273 were being counted as actual protocol errors, and logged if the log
7274 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7275 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7276 and DATA following a set of rejected RCPTs do not count as protocol errors.
7277 In other words, Exim assumes they were pipelined, though this may not
7278 actually be the case. Of course, in all cases the client gets an
7279 appropriate error code.
7281 15. If a lookup fails in an ACL condition, a message about the failure may
7282 be available; it is used if testing the ACL cannot continue, because most
7283 such messages specify what the cause of the deferral is. However, some
7284 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7285 that caused an old message to be retained and used if a later statement
7286 caused a defer, replacing the real cause of the deferral.
7288 16. If an IP address had so many PTR records that the DNS lookup buffer
7289 was not large enough to hold them, Exim could crash while trying to process
7290 the truncated data. It now detects and logs this case.
7292 17. Further to 4.21/58, another change has been made: if (and only if) the
7293 first line of a message (the first header line) ends with CRLF, a bare LF
7294 in a subsequent header line has a space inserted after it, so as not to
7295 terminate the header.
7297 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7298 unnecessarily, used atoi() instead of strtol(), and didn't check the
7299 termination when getting file sizes from file names by regex.
7301 19. Completely re-implemented the support for maildirsize files, in the light
7302 of a number of problems with the previous contributed implementation
7303 (4.30/29). In particular:
7305 . If the quota is zero, the maildirsize file is maintained, but no quota is
7308 . If the maildir directory does not exist, it is created before any attempt
7309 to write a maildirsize file.
7311 . The quota value in the file is just a cache; if the quota is changed in
7312 the transport, the new value overrides.
7314 . A regular expression is available for excluding directories from the
7317 20. The autoreply transport checks the characters in options that define the
7318 message's headers; it allows continued headers, but it was checking with
7319 isspace() after an embedded newline instead of explicitly looking for a
7322 21. If all the "regular" hosts to which an address was routed had passed their
7323 expiry times, and had not reached their retry times, the address was
7324 bounced, even if fallback hosts were defined. Now Exim should go on to try
7327 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7328 equivalent code in the SMTP transport. Some hosts send humungous responses
7329 to HELO/EHLO, more than 1024 it seems.
7331 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7332 gives compiler warnings in some environments. I've now done it "properly",
7335 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7336 (because of problems with the built-in one) was declared to return uschar *
7337 instead of char *, causing compiler failure.
7339 25. Fixed a file descriptor leak when processing alias/forward files.
7341 26. Fixed a minor format string issue in dbfn.c.
7343 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7345 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7346 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7349 29. When a sender address is verified, it is cached, to save repeating the test
7350 when there is more than one recipient in a message. However, when the
7351 verification involves a callout, it is possible for different callout
7352 options to be set for different recipients. It is too complicated to keep
7353 track of this in the cache, so now Exim always runs a verification when a
7354 callout is required, relying on the callout cache for the optimization.
7355 The overhead is duplication of the address routing, but this should not be
7358 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7359 to be verified with callout=postmaster, and the main callout worked but the
7360 postmaster check failed, the verification correctly failed. However, if a
7361 subsequent RCPT command asked for sender verification *without* the
7362 postmaster check, incorrect caching caused this verification also to fail,
7365 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7366 it was not caching the DNS options (qualify_single, search_parents) that
7367 were used when the lookup failed. A subsequent lookup with different
7368 options therefore always gave the same answer, though there were cases
7369 where it should not have. (Example: a "domains = !$mx_any" option on a
7370 dnslookup router: the "domains" option is always processed without any
7371 widening, but the router might have qualify_single set.) Now Exim uses the
7372 cached value only when the same options are set.
7374 32. Added John Jetmore's "exipick" utility to the distribution.
7376 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7377 than a timeout (e.g. a certificate is required, and is not provided), an
7378 Exim server now closes the connection immediately. Previously it waited for
7379 the client to close - but if the client is SSL, it seems that they each
7380 wait for each other, leading to a delay before one of them times out.
7382 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7383 maintained 0.8.x compatibility because I don't think many are using it, and
7384 it is clearly obsolete.
7386 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7389 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7390 client certificate was expired. A simple patch fixes this, though I don't
7391 understand the full logic of why the verify callback is called multiple
7394 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7395 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7396 which causes problems with some clients (such as the Certicom SSL Plus
7397 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7398 disables the coutermeasure allowing Eudora to connect."
7400 38. Exim was not checking that a write() to a log file succeeded. This could
7401 lead to Bad Things if a log got too big, in particular if it hit a file
7402 size limit. Exim now panics and dies if it cannot write to a log file, just
7403 as it does if it cannot open a log file.
7405 39. Modified OS/Makefile-Linux so that it now contains
7407 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7409 The two -D definitions ensure that Exim is compiled with large file
7410 support, which makes it possible to handle log files that are bigger than
7413 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7414 instance) a domain was checked against a named list that involved a lookup,
7415 causing $domain_data to be set, then another domain was checked against the
7416 same list, then the first domain was re-checked, the value of $domain_data
7417 after the final check could be wrong. In particular, if the second check
7418 failed, it could be set empty. This bug probably also applied to
7421 41. The strip_trailing_dot option was not being applied to the address given
7422 with the -f command-line option.
7424 42. The code for reading a message's header from the spool was incrementing
7425 $received_count, but never initializing it. This meant that the value was
7426 incorrect (doubled) while delivering a message in the same process in which
7427 it was received. In the most common configuration of Exim, this never
7428 happens - a fresh exec is done - but it can happen when
7429 deliver_drop_privilege is set.
7431 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7432 it now includes up to 150 characters of the unexpected data in the log
7435 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7436 and building data strings. The size of both of these buffers was 10 000
7437 bytes - far larger than anybody would *ever* want, thought I. Needless to
7438 say, somebody hit the limit. I have increased the maximum line length to
7439 20 000 and the maximum data length of concatenated lines to 100 000. I have
7440 also fixed two bugs, because there was no checking on these buffers. Tsk,
7441 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7442 buffer is too small.
7444 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7445 lsearch lookups. Now it does.
7447 46. When parsing a route_list item in a manualroute router, a fixed-length
7448 buffer was used for the list of hosts. I made this 1024 bytes long,
7449 thinking that nobody would ever have a list of hosts that long. Wrong.
7450 Somebody had a whole pile of complicated expansion conditions, and the
7451 string was silently truncated, leading to an expansion error. It turns out
7452 that it is easier to change to an unlimited length (owing to other changes
7453 that have happened since this code was originally written) than to build
7454 structure for giving a limitation error. The length of the item that
7455 expands into the list of hosts is now unlimited.
7457 47. The lsearch lookup could not handle data where the length of text line was
7458 more than 4095 characters. Such lines were truncated, leading to shortened
7459 data being returned. It should now handle lines of any length.
7461 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7462 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7465 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7466 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7467 no longer happens in comment lines. A list of replacements is now placed
7468 at the head of all of the source files, except those whose only change is
7469 to replace PERL_COMMAND in the very first #! line.
7471 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7472 messages on the queue, with a bottom-up merge sort, using code contributed
7473 by Michael Haardt. This should make operations like -bp somewhat faster on
7474 large queues. It won't affect queue runners, except when queue_run_in_order
7477 51. Installed eximstats 1.31 in the distribution.
7479 52. Added support for SRV lookups to the dnslookup router.
7481 53. If an ACL referred to $message_body or $message_body_end, the value was not
7482 reset for any messages that followed in the same SMTP session.
7484 54. The store-handling optimization for building very long strings was not
7485 differentiating between the different store pools. I don't think this
7486 actually made any difference in practice, but I've tidied it.
7488 55. While running the routers to verify a sender address, $sender_address
7489 was still set to the sender address. This is wrong, because when routing to
7490 send a bounce to the sender, it would be empty. Therefore, I have changed
7491 it so that, while verifying a sender address, $sender_address is set to <>.
7492 (There is no change to what happens when verifying a recipient address.)
7494 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7495 target A or AAAA records (if not already returned) without resetting the
7496 qualify_single or search_parents options of the DNS resolver. These are
7497 inappropriate in this case because the targets of MX and SRV records must
7498 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7499 target that, when qualified, matched something in the local domain. These
7500 two options are now turned off when doing these lookups.
7502 57. It seems that at least some releases of Reiserfs (which does not have the
7503 concept of a fixed number of inodes) returns zero and not -1 for the
7504 number of available inodes. This interacted badly with check_spool_inodes,
7505 which assumed that -1 was the "no such thing" setting. What I have done is
7506 to check that the total number of inodes is greater than zero before doing
7507 the test of how many are available.
7509 58. When a "warn" ACL statement has a log_message modifier, the message is
7510 remembered, and not repeated. This is to avoid a lot of repetition when a
7511 message has many recipients that cause the same warning to be written.
7512 However, Exim was preserving the list of already written lines for an
7513 entire SMTP session, which doesn't seem right. The memory is now reset if a
7514 new message is started.
7516 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7517 debugging output unless log_rewrite was also set.
7519 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7520 of (int)(handle) when we know that handle contains (void *)(-1).
7522 61. The Exim daemon panic-logs an error return when it closes the incoming
7523 connection. However "connection reset by peer" seems to be common, and
7524 isn't really an error worthy of noting specially, so that particular error
7527 62. When Exim is trying to find all the local interfaces, it used to panic and
7528 die if the ioctl to get the interface flags failed. However, it seems that
7529 on at least one OS (Solaris 9) it is possible to have an interface that is
7530 included in the list of interfaces, but for which you get a failure error
7531 for this call. This happens when the interface is not "plumbed" into a
7532 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7533 failure of the "get flags" call assumes that the interface is down.
7535 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7536 makes life easier for people who are doing arithmetic on fields extracted
7537 from dates, where you often get leading zeros that should not be
7538 interpreted as octal.
7540 64. Added qualify_domain to the redirect router, to override the global
7543 65. If a pathologically long header line contained very many addresses (the
7544 report of this problem mentioned 10 000) and each of them was rewritten,
7545 Exim could use up a very large amount of memory. (It kept on making new
7546 copies of the header line as it rewrote, and never released the old ones.)
7547 At the expense of a bit more processing, the header rewriting function has
7548 been changed so that it no longer eats memory in this way.
7550 66. The generation of the Received: header has been moved from the time that a
7551 message starts to be received, to the time that it finishes. The timestamp
7552 in the Received: header should now be very close to that of the <= log
7553 line. There are two side-effects of this change:
7555 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7556 logged header lines no longer include the local Received: line, because
7557 it has not yet been created. The same applies to a copy of the message
7558 that is returned to a non-SMTP sender when a message is rejected.
7560 (b) When a filter file is tested using -bf, no additional Received: header
7561 is added to the test message. After some thought, I decided that this
7564 This change does not affect the value of $received_for. It is still set
7565 after address rewriting, but before local_scan() is called.
7567 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7569 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7570 gave an unhelpful panic error message, and a defer error. I have managed to
7571 change this behaviour so that it now rejects any supplied certificate,
7572 which seems right, as the list of acceptable certificates is empty.
7574 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7575 gave an unhelpful defer error. I have not managed to make this reject any
7576 supplied certificates, but the error message it gives is "no certificate
7577 supplied", which is not helpful.
7579 70. exigrep's output now also includes lines that are not associated with any
7580 message, but which match the given pattern. Implemented by a patch from
7581 Martin Sluka, which also tidied up the Perl a bit.
7583 71. Recipient callout verification, like sender verification, was using <> in
7584 the MAIL FROM command. This isn't really the right thing, since the actual
7585 sender may affect whether the remote host accepts the recipient or not. I
7586 have changed it to use the actual sender in the callout; this means that
7587 the cache record is now keyed on a recipient/sender pair, not just the
7588 recipient address. There doesn't seem to be a real danger of callout loops,
7589 since a callout by the remote host to check the sender would use <>.
7590 [SEE ABOVE: changed after hitting problems.]
7592 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7593 temporary errors. However, in the case of such a code being given after
7594 the end of a data transmission (i.e. after ".") Exim was failing to write
7595 a retry record for the message. (Yes, there was some broken host that was
7596 actually sending 8xx at this point.)
7598 73. An unknown lookup type in a host list could cause Exim to panic-die when
7599 the list was checked. (An example that provoked this was putting <; in the
7600 middle of a list instead of at the start.) If this happened during a DATA
7601 ACL check, a -D file could be left lying around. This kind of configuration
7602 error no longer causes Exim to die; instead it causes a defer error. The
7603 incident is still logged to the main and panic logs.
7605 74. Buglet left over from Exim 3 conversion. The message "too many messages
7606 in one connection" was written to the rejectlog but not the mainlog, except
7607 when address rewriting (yes!) was being logged.
7609 75. Added write_rejectlog option.
7611 76. When a system filter was run not as root (that is, when system_filter_user
7612 was set), the values of the $n variables were not being returned to the
7613 main process; thus, they were not subsequently available in the $sn
7616 77. Added +return_path_on_delivery log selector.
7618 78. A connection timeout was being treated differently from recipients deferred
7619 when testing hosts_max_try with a message that was older than the host's
7620 retry timeout. (The host should not be counted, thus allowing all hosts to
7621 be tried at least once before bouncing.) This may have been the cause of an
7622 occasionally reported bug whereby a message would remain on the queue
7623 longer than the retry timeout, but would be bounced if a delivery was
7624 forced. I say "may" because I never totally pinned down the problem;
7625 setting up timeout/retry tests is difficult. See also the next item.
7627 79. The ultimate address timeout was not being applied to errors that involved
7628 a combination of host plus message (for example, a timeout on a MAIL
7629 command). When an address resolved to a number of possible hosts, and they
7630 were not all tried for each delivery (e.g. because of hosts_max_try), a
7631 message could remain on the queue longer than the retry timeout.
7633 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7636 81. Fixed an obscure SMTP outgoing bug which required at least the following
7637 conditions: (a) there was another message waiting for the same server;
7638 (b) the server returned 5xx to all RCPT commands in the first message so
7639 that the message was not completed; (c) the server dropped the connection
7640 or gave a negative response to the RSET that Exim sends to abort the
7641 transaction. The observed case was a dropped connection after DATA that had
7642 been sent in pipelining mode. That is, the server had advertised PIPELINING
7643 but was not implementing it correctly. The effect of the bug was incorrect
7644 behaviour, such as trying another host, and this could lead to a crash.
7650 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7651 and daemon.c were passed as pointers to ints; they should have been
7652 pointers to socklen_t variables (which are typically unsigned ints).
7654 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7657 3. Fixed a really odd bug that affected only the testing scheme; patching a
7658 certain fixed string in the binary changed the value of another string that
7659 happened to be identical to the end of the original first string.
7661 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7662 name", it returns that address as the IP address. On some operating
7663 systems (e.g. Solaris), it also passes back the IP address string as the
7664 "host name". However, on others (e.g. Linux), it passes back an empty
7665 string. Exim wasn't checking for this, and was changing the host name to an
7666 empty string, assuming it had been canonicalized.
7668 5. Although rare, it is permitted to have more than one PTR record for a given
7669 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7670 all the names associated with an address, because they do in Solaris.
7671 However, it seems that they do not in Linux for data that comes from the
7672 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7673 I found this out when I moved to a new Linux workstation and tried to run
7674 the Exim test suite.
7676 To get round this problem I have changed the code so that it now does its
7677 own call to the DNS to look up PTR records when searching for a host name.
7678 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7679 addresses that are only in /etc/hosts are still found.
7681 This behaviour is, however, controlled by an option called host_lookup_
7682 order, which defaults to "bydns:byaddr". If people want to use the other
7683 order, or indeed, just use one or the other means of lookup, they can
7684 specify it in this variable.
7686 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7687 some operating systems, this comes back from gethostbyaddr() as an empty
7688 string, and this is what Exim used to test for. However, it seems that in
7689 other systems, "." is yielded. Exim now tests for this case too.
7691 7. The values of check_spool_space and check_log_space are now held internally
7692 as a number of kilobytes instead of an absolute number of bytes. If a
7693 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7694 kilobyte. This means that much larger values can be stored.
7696 8. Exim monitor: an attempt to get the action menu when not actually pointing
7697 at a message produces an empty menu entitled "No message selected". This
7698 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7699 no entries in it ("Shell widget menu has zero width and/or height"). So I
7700 have added a single, blank menu entry in this case.
7702 9. Added ${quote_local_part.
7704 10. MIME decoding is now applied to the contents of Subject: header lines when
7707 11. Now that a reference to $sender_host_address automatically causes a reverse
7708 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7709 host lookup before query-style lookups in lists that might use this
7710 variable. This has therefore been abolished, and the "net-" prefix is no
7711 longer necessary for query-style lookups.
7713 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7714 have been a typo for LFLAGS, so it has been changed.
7716 13. The install script calls Exim with "-C /dev/null" in order to find the
7717 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7718 to be output. However, since Exim outputs its version number before the
7719 error, it didn't break the script. It just looked ugly. I fixed this by
7720 always allowing "-C /dev/null" if the caller is root.
7722 14. Ignore overlarge ACL variable number when reading spool file - insurance
7723 against a later release with more variables having written the file.
7725 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7726 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7727 this, as well as the form without the "IPv6" on the front.
7729 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7730 OS/Makefile-Darwin file.
7732 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7733 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7735 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7736 variable. However, some OS use ints for inodes. I've added cast to long int
7737 to get rid of the compiler warning.
7739 19. I had forgotten to lock out "/../" in configuration file names when
7740 ALT_CONFIG_PREFIX was set.
7742 20. Routers used for verification do not need to specify transports. However,
7743 if such a router generated a host list, and callout was configured, Exim
7744 crashed, because it could not find a port number from the (non-existent)
7745 transport. It now assumes port 25 in this circumstance.
7747 21. Added the -t option to exigrep.
7749 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7750 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7751 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7752 them is set, LOOKUP_LSEARCH is forced.
7754 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7756 24. Added sender and host information to the "rejected by local_scan()" log
7757 line; previously there was no indication of these.
7759 25. Added .include_if_exists.
7761 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7762 Linux. It turns out that not all file systems support this. Apparently some
7763 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7764 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7765 not supported on the file descriptor, is now ignored when Exim is trying to
7766 sync a directory. This applies only to Linux.
7768 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7770 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7771 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7774 29. Added support for maildirsize files from supplied patch (modified a bit).
7776 30. The use of :fail: followed by an empty string could lead Exim to respond to
7777 sender verification failures with (e.g.):
7779 550 Verification failed for <xxx>
7780 550 Sender verify failed
7782 where the first response line was missing the '-' that indicates it is not
7783 the final line of the response.
7785 31. The loop for finding the name of the user that called Exim had a hardwired
7786 limit of 10; it now uses the value of finduser_retries, which is used for
7787 all other user lookups.
7789 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7792 33. Exim was neglecting to zero errno before one call of strtol() when
7793 expanding a string and expecting an integer value. On some systems this
7794 resulted in spurious "integer overflow" errors. Also, it was casting the
7795 result into an int without checking.
7797 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7798 did not work. The code looks as if it has *never* worked, though it appears
7799 to have been documented since at least release 1.62. I have made it work.
7801 35. The "timeout_DNS" error in retry rules, also documented since at least
7802 1.62, also never worked. As it isn't clear exactly what this means, and
7803 clearly it isn't a major issue, I have abolished the feature by treating it
7804 as "timeout", and writing a warning to the main and panic logs.
7806 36. The display of retry rules for -brt wasn't always showing the error code
7809 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7810 timeout_connect_A, timeout_connect_MX.
7812 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7813 to the empty sender.
7815 39. The daemon was not analysing the content of -oX till after it had closed
7816 stderr and disconnected from the controlling terminal. This meant that any
7817 syntax errors were only noted on the panic log, and the return code from
7818 the command was 0. By re-arranging the code a little, I've made the
7819 decoding happen first, so such errors now appear on stderr, and the return
7820 code is 1. However, the actual setting up of the sockets still happens in
7821 the disconnected process, so errors there are still only recorded on the
7824 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7825 connections (as happens on some IP stacks) was logged at start up time as
7826 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7827 it from "IPv6 and IPv4", which means that two separate sockets are being
7830 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7831 says whether AF_INET or AF_INET6 was passed as an argument.
7833 42. Exiwhat output was messed up when time zones were included in log
7836 43. Exiwhat now gives more information about the daemon's listening ports,
7837 and whether -tls-on-connect was used.
7839 44. The "port" option of the smtp transport is now expanded.
7841 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7842 silently ignored. Now an error message is written to the main and panic
7845 46. There's a new ACL modifier called "logwrite" which writes to a log file
7846 as soon as it is encountered.
7848 47. Added $local_user_uid and $local_user_gid at routing time.
7850 48. Exim crashed when trying to verify a sender address that was being
7853 49. Exim was recognizing only a space character after ".include". It now also
7854 recognizes a tab character.
7856 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7857 extracting the relevant information from the specification. The man page no
7858 longer contains scrambled data for the -d option, and I've added a section
7859 at the front about calling Exim under different names.
7861 51. Added "extra_headers" argument to the "mail" command in filter files.
7863 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7866 53. Installed eximstats 1.29.
7868 54. Added transport_filter_timeout as a generic transport option.
7870 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7871 Cc: header lines. This was required by RFC 822, but it not required by RFC
7874 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7875 incoming messages that did not have them. Now it does so only if the
7876 message originates locally, that is, if there is no associated remote host
7877 address. When Resent- header lines are present, this applies to the Resent-
7878 lines rather than the non-Resent- lines.
7880 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7881 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7883 58. Messages for configuration errors now include the name of the main
7884 configuration file - useful now that there may be more than one file in a
7885 list (.included file names were always shown).
7887 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7888 for those rare installations that do not start the daemon as root or run it
7889 setuid root. I've cut out the call to initgroups() if the daemon is not
7892 60. The Exim user and group can now be bound into the binary as text strings
7893 that are looked up at the start of Exim's processing.
7895 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7897 62. Added $mailstore_basename variable.
7899 63. Installed patch to sieve.c from Michael Haardt.
7901 64. When Exim failed to open the panic log after failing to open the main log,
7902 the original message it was trying to log was written to stderr and debug
7903 output, but if they were not available (the usual case in production), it
7904 was lost. Now it is written to syslog before the two lines that record the
7905 failures to open the logs.
7907 65. Users' Exim filters run in subprocesses under the user's uid. It is
7908 possible for a "deliver" command or an alias in a "personal" command to
7909 provoke an address rewrite. If logging of address rewriting is configured,
7910 this fails because the process is not running as root or exim. There may be
7911 a better way of dealing with this, but for the moment (because 4.30 needs
7912 to be released), I have disabled address rewrite logging when running a
7913 filter in a non-root, non-exim process.
7919 1. The buildconfig auxiliary program wasn't quoting the value set for
7920 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7921 not defined. This bug was masked in 4.22 by the effect that was fixed in
7924 2. Some messages that were rejected after a message id was allocated were
7925 shown as "incomplete" by exigrep. It no longer does this for messages that
7926 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7928 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7929 have allow_domain_literals set, the ID did not get logged in the <= line.
7930 Domain literals are now always recognized in Message-ID: header lines.
7932 4. The first argument for a ${extract expansion item is the key name or field
7933 number. Leading and trailing spaces in this item were not being ignored,
7934 causing some misleading effects.
7936 5. When deliver_drop_privilege was set, single queue runner processes started
7937 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7938 same command in the process it spins off) were not dropping privilege.
7940 6. When the daemon running as "exim" started a queue runner, it always
7941 re-executed Exim in the spun-off process. This is a waste of effort when
7942 deliver_drop_privilege is set. The new process now just calls the
7943 queue-runner function directly.
7949 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7952 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7953 ignored. Though the use of -oP was forcing the writing of a pid file, it
7954 was always written to the default place.
7956 3. If the message "no IP address found for host xxxx" is generated during
7957 incoming verification, it is now followed by identification of the incoming
7958 connection (so you can more easily find what provoked it).
7960 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7962 5. Added some features to "harden" Exim a bit more against certain attacks:
7964 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7965 be put in Local/Makefile. This is like the never_users runtime option,
7966 but it cannot be overridden. The default setting is "root".
7968 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7969 prefix string with which any file named in a -C command line option
7972 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7973 is retained for -C and -D only if the caller of Exim is root. Without
7974 it, the exim user may also use -C and -D and retain privilege.
7976 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7977 command line option is disabled.
7979 6. Macro names set by the -D option must start with an upper case letter, just
7980 like macro names defined in the configuration file.
7982 7. Added "dereference=" facility to LDAP.
7984 8. Two instances of the typo "uknown" in the source files are fixed.
7986 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7987 the Configure-Makefile script screwed up while processing it.
7989 10. Incorporated PCRE 4.4.
7991 11. The SMTP synchronization check was not operating right at the start of an
7992 SMTP session. For example, it could not catch a HELO sent before the client
7993 waited for the greeting. There is now a check for outstanding input at the
7994 point when the greeting is written. Because of the duplex, asynchronous
7995 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7996 way, but not yet received, when the check is performed.
7998 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7999 on TCP/IP sockets, because this apparently causes some broken clients to
8002 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8003 unchanged) from the Cygwin maintainer.
8005 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8006 was supported instead of testing for mbx. Effectively a typo.
8008 15. The spa authenticator server code was not checking that the input it
8009 received was valid base64.
8011 16. The debug output line for the "set" modifier in ACLs was not showing the
8012 name of the variable that was being set.
8014 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8016 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8017 lookup on its own. Something else was needed to trigger the lookup. For
8018 example, a match in host_lookup or the need for a host name in a host list.
8019 Now, if $sender_host_name is referenced and the host name has not yet been
8020 looked up, a lookup is performed. If the lookup fails, the variable remains
8021 empty, and $host_lookup_failed is set to "1".
8023 19. Added "eqi" as a case-independent comparison operator.
8025 20. The saslauthd authentication condition could segfault if neither service
8026 nor realm was specified.
8028 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8029 error message that was logged was misleading, and incoming SMTP
8030 connections were dropped. The message is now more accurate, and temporary
8031 errors are given to SMTP connections.
8033 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8034 (except RSET) with a 421 error, until QUIT is received. However, it was
8035 failing to send a response to QUIT.
8037 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8038 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8039 ACL is now run after the helo_try_verify_hosts code.
8041 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8042 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8043 case-independent, so other case variants are also recognized). Apparently
8044 some systems use these upper case variants.
8046 25. If more than two messages were waiting for the same host, and a transport
8047 filter was specified for the transport, Exim sent two messages over the
8048 same TCP/IP connection, and then failed with "socket operation on non-
8049 socket" when it tried to send the third.
8051 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8053 27. The extern definition of crypt16() in expand.c was not being excluded when
8054 the OS had its own crypt16() function.
8056 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8057 as a preferred synonym for return_size_limit, both as an option and as an
8060 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8062 30. Changed the default configuration ACL to relax the local part checking rule
8063 for addresses that are not in any local domains. For these addresses,
8064 slashes and pipe symbols are allowed within local parts, but the sequence
8065 /../ is explicitly forbidden.
8067 31. SPA server authentication was not clearing the challenge buffer before
8070 32. log_message in a "warn" ACL statement was writing to the reject log as
8071 well as to the main log, which contradicts the documentation and doesn't
8072 seem right (because no rejection is happening). So I have stopped it.
8074 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8075 However, I am unable to do any testing of this.
8077 34. Fixed an infelicity in the appendfile transport. When checking directories
8078 for a mailbox, to see if any needed to be created, it was accidentally
8079 using path names with one or more superfluous leading slashes; tracing
8080 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8082 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8083 contents are added to the log line that is written for every discarded
8084 recipient. (Previously a log_message setting was ignored.)
8086 36. The ${quote: operator now quotes the string if it is empty.
8088 37. The install script runs exim in order to find its version number. If for
8089 some reason other than non-existence or emptiness, which it checks, it
8090 could not run './exim', it was installing it with an empty version number,
8091 i.e. as "exim-". This error state is now caught, and the installation is
8094 38. An argument was missing from the function that creates an error message
8095 when Exim fails to connect to the socket for saslauthd authentication.
8096 This could cause Exim to crash, or give a corrupted message.
8098 39. Added isip, isip4, and isip6 to ${if conditions.
8100 40. The ACL variables $acl_xx are now saved with the message, and can be
8101 accessed later in routers, transports, and filters.
8103 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8104 strings in the file are not string-expanded.
8106 42. If a MAIL command specified a SIZE value that was too large to fit into an
8107 int variable, the check against message_size_limit failed. Such values are
8108 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8109 day this will have to be increased, but I don't think I want to be around
8110 when emails are that large.
8117 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8118 iconv() is not standard in FreeBSD.
8120 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8121 IPv6 enabled. The observed symptom was a segmentation fault on return from
8122 the function os_common_find_running_interfaces() in src/os.c.
8124 3. In the check_special_case() function in daemon.c I had used "errno" as an
8125 argument name, which causes warnings on some systems. This was basically a
8126 typo, since it was named "eno" in the comments!
8128 4. The code that waits for the clock to tick (at a resolution of some fraction
8129 of a second) so as to ensure message-id uniqueness was always waiting for
8130 at least one whole tick, when it could have waited for less. [This is
8131 almost certainly not relevant at current processor speeds, where it is
8132 unlikely to ever wait at all. But we try to future-proof.]
8134 5. The function that sleeps for a time interval that includes fractions of a
8135 second contained a race. It did not block SIGALRM between setting the
8136 timer, and suspending (a couple of lines later). If the interval was short
8137 and the sigsuspend() was delayed until after it had expired, the suspension
8138 never ended. On busy systems this could lead to processes getting stuck for
8141 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8142 process, before it forks any delivery processes. The open lookup caching
8143 mechanism meant that the open file or database connection was passed into
8144 the delivery process. The problem was that delivery processes always tidy
8145 up cached lookup data. This could cause a problem for the next delivery
8146 process started by the queue runner, because the external queue runner
8147 process does not know about the closure. So the next delivery process
8148 still has data in the lookup cache. In the case of a file lookup, there was
8149 no problem because closing a file descriptor in a subprocess doesn't affect
8150 the parent. However, if the lookup was caching a connection to a database,
8151 the connection was closed, and the second delivery process was likely to
8152 see errors such as "PGSQL: query failed: server closed the connection
8153 unexpectedly". The problem has been fixed by closing all cached lookups
8154 in a queue runner before running a delivery process.
8156 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8157 seem to have the "const" qualifier which it has on other OS. I've
8160 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8161 *of the same type* (client or server) with the same public name that an
8162 error should be diagnosed.
8164 9. When Exim looked up a host name for an IP address, but failed to find the
8165 original IP address when looking up the host name (a safety check), it
8166 output the message "<ip address> does not match any IP for NULL", which was
8167 confusing, to say the least. The bug was that the host name should have
8168 appeared instead of "NULL".
8170 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8171 user that is built into the binary, and the -C or -D options is used, root
8172 privilege is dropped before the configuration file is read. In addition,
8173 logging is switched to stderr instead of the normal log files. If the
8174 configuration then re-defines the Exim user, the unprivileged environment
8175 is probably not what is expected, so Exim logs a panic warning message (but
8178 However, if deliver_drop_privilege is set, the unprivileged state may well
8179 be exactly what is intended, so the warning has been cut out in that case,
8180 and Exim is allowed to try to write to its normal log files.
8186 1. smtp_return_error_details was not giving details for temporary sender
8187 or receiver verification errors.
8189 2. Diagnose a configuration error if two authenticators have the same public
8192 3. Exim used not to create the message log file for a message until the first
8193 delivery attempt. This could be confusing when incoming messages were held
8194 for policy or load reasons. The message log file is now created at the time
8195 the message is received, and an initial "Received" line is written to it.
8197 4. The automatically generated man page for command line options had a minor
8198 bug that caused no ill effects; however, a more serious problem was that
8199 the procedure for building the man page automatically didn't always
8200 operate. Consequently, release 4.20 contains an out-of-date version. This
8201 shouldn't happen again.
8203 5. When building Exim with embedded Perl support, the script that builds the
8204 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8205 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8207 6. The freeze_tell option was not being used for messages that were frozen on
8208 arrival, either by an ACL or by local_scan().
8210 7. Added the smtp_incomplete_transaction log selector.
8212 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8213 was accepting AUTH without a new EHLO.
8215 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8216 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8219 10. Logging of TCP/IP connections (when configured) now happens in the main
8220 daemon process instead of the child process, so that the TCP/IP connection
8221 count is more accurate (but it can never be perfect).
8223 11. The use of "drop" in a nested ACL was not being handled correctly in the
8224 outer ACL. Now, if condition failure induced by the nested "drop" causes
8225 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8226 or "require"), the connection is dropped.
8228 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8229 that yield "discard" can now be used with an "accept" or a "discard" verb,
8230 but an error is generated for any others (because I can't see a useful way
8231 to define what should happen).
8233 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8234 are now processed in the same way as lines in the Exim configuration file.
8235 In particular, continuation lines are supported.
8237 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8239 15. Added -ti meaning -t -i.
8241 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8242 domains, and warn by logging if others are found.
8244 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8245 returned from GIFCONF when Exim is trying to find the list of interfaces on
8246 a host. The code in os.c has been modified to copy each ifreq to an aligned
8247 structure in all cases.
8249 Also, in some cases, the returned ifreq's were being copied to a 'struct
8250 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8251 means the last couple of bytes of an IPv6 address could be chopped if the
8252 ifreq contained only a normal sockaddr (14 bytes storage).
8254 18. Named domain lists were not supported in the hosts_treat_as_local option.
8255 An entry such as +xxxx was not recognized, and was treated as a literal
8258 19. Ensure that header lines added by a DATA ACL are included in the reject log
8259 if the ACL subsequently rejects the message.
8261 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8262 MD5 (which is deprecated).
8264 21. When testing a filter file using -bf, Exim was writing a message when it
8265 took the sender from a "From " line in the message, but it was not doing so
8266 when it took $return_path from a Return-Path: header line. It now does.
8268 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8269 with a valid header line field name (a series of printing characters
8270 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8272 23. Changed "disc" in the source to "disk" to conform to the documentation and
8273 the book and for uniformity.
8275 24. Ignore Sendmail's -Ooption=value command line item.
8277 25. When execve() failed while trying to run a command in a pipe transport,
8278 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8279 could be confused with a return value of 69 from the command itself. This
8280 has been changed to 127, the value the shell returns if it is asked to run
8281 a non-existent command. The wording for the related log line suggests a
8282 non-existent command as the problem.
8284 26. If received_header_text expands to an empty string, do not add a Received:
8285 header line to the message. (Well, it adds a token one on the spool, but
8286 marks it "old" so that it doesn't get used or transmitted.)
8288 27. Installed eximstats 1.28 (addition of -nt option).
8290 28. There was no check for failure on the call to getsockname() in the daemon
8291 code. This can fail if there is a shortage of resources on the system, with
8292 ENOMEM, for example. A temporary error is now given on failure.
8294 29. Contrary to the C standard, it seems that in some environments, the
8295 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8296 program. Exim now does this explicitly; it affects the formatting of
8297 timestamps using strftime().
8299 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8300 complaints. I've now initialized all the variables, to avoid this.
8302 32. Header lines added by a system filter were not being "seen" during
8303 transport-time rewrites.
8305 33. The info_callback() function passed to OpenSSL is set up with type void
8306 (*)(SSL *, int, int), as described somewhere. However, when calling the
8307 function (actually a macro) that sets it up, the type void(*)() is
8308 expected. I've put in a cast to prevent warnings from picky compilers.
8310 34. If a DNS black list lookup found a CNAME record, but there were no A
8311 records associated with the domain it pointed at, Exim crashed.
8313 35. If a DNS black list lookup returned more than one A record, Exim ignored
8314 all but the first. It now scans all returned addresses if a particular IP
8315 value is being sought. In this situation, the contents of the
8316 $dnslist_value variable are a list of all the addresses, separated by a
8319 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8320 to accept a host name and all its aliases if the forward lookup for any of
8321 them yielded the IP address of the incoming connection. Now it accepts only
8322 those names whose forward lookup yields the correct IP address. Any other
8323 names are discarded. This closes a loophole whereby a rogue DNS
8324 administrator could create reverse DNS records to break through a
8325 wildcarded host restriction in an ACL.
8327 37. If a user filter or a system filter that ran in a subprocess used any of
8328 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8329 the wrong values were passed to the pipe command ($thisaddress had the
8330 value of $0, $0 had the value of $1, etc). This bug was introduced by
8331 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8333 38. Improved the line breaking for long SMTP error messages from ACLs.
8334 Previously, if there was no break point between 40 and 75 characters, Exim
8335 left the rest of the message alone. Two changes have been made: (a) I've
8336 reduced the minimum length to 35 characters; (b) if it can't find a break
8337 point between 35 and 75 characters, it looks ahead and uses the first one
8338 that it finds. This may give the occasional overlong line, but at least the
8339 remaining text gets split now.
8341 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8342 file descriptors might be low, and that setting 1000 would always raise it.
8343 It turns out that in some environments, the limit is already over 1000 and
8344 that lowering it causes trouble. So now Exim takes care not to decrease it.
8346 40. When delivering a message, the value of $return_path is set to $sender_
8347 address at the start of routing (routers may change the value). By an
8348 oversight, this default was not being set up when an address was tested by
8349 -bt or -bv, which affected the outcome if any router or filter referred to
8352 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8353 log only when "message" or "log_message" are set. However, if one of the
8354 conditions was an address verification, or a call to a nested ACL, the
8355 messages generated by the underlying test were being passed through. This
8356 no longer happens. The underlying message is available in $acl_verify_
8357 message for both "message" and "log_message" expansions, so it can be
8358 passed through if needed.
8360 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8361 new expansion $bh_ to give the encoded byte string without charset
8362 translation. Translation happens only if iconv() is available; HAVE_ICONV
8363 indicates this at build time. HEADERS_CHARSET gives the charset to
8364 translate to; headers_charset can change it in the configuration, and
8365 "headers charset" can change it in an individual filter file.
8367 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8368 that creates RFC 2047 encoded "words" labels them as that charset instead
8369 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8370 expansion operator; (ii) when Exim creates a From: line for a local
8371 message; (iii) when a header line is rewritten to include a "phrase" part.
8373 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8374 buggy, causing it to skip the first lines of messages whose message ID
8375 ended in 'D'. This would not have bitten before Exim release 4.14, because
8376 message IDs were unlikely to end in 'D' before then. The effect was to have
8377 incorrect size information for certain domains.
8379 45. #include "config.h" was missing at the start of the crypt16.c module. This
8380 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8383 46. If there was a timeout during a "random" callout check, Exim treated it as
8384 a failure of the random address, and carried on sending RSET and the real
8385 address. If the delay was just some slowness somewhere, the response to the
8386 original RCPT would be taken as a response to RSET and so on, causing
8387 mayhem of various kinds.
8389 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8390 when I implemented it. It didn't allow for the fact that some option values
8391 may legitimately be negative (e.g. size_addition), and it didn't even do
8392 the right test for positive values.
8394 48. Domain names in DNS records are case-independent. Exim always looks them up
8395 in lower case. Some resolvers return domain names in exactly the case they
8396 appear in the zone file, that is, they may contain uppercase letters. Not
8397 all resolvers do this - some return always lower case. Exim was treating a
8398 change of case by a resolver as a change of domain, similar to a widening
8399 of a domain abbreviation. This triggered its re-routing code and so it was
8400 trying to route what was effectively the same domain again. This normally
8401 caused routing to fail (because the router wouldn't handle the domain
8402 twice). Now Exim checks for this case specially, and just changes the
8403 casing of the domain that it ultimately uses when it transmits the message
8406 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8409 50. If a filter generated a file delivery with a non-absolute name (possible if
8410 no home directory exists for the router), the forbid_file option was not
8413 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8414 the existing equality matching.
8416 52. Exim was using ints instead of ino_t variables in some places where it was
8417 dealing with inode numbers.
8419 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8420 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8421 called TMPDIR, and if it finds it is different, it changes its value.
8423 54. The smtp_printf() function is now made available to local_scan() so
8424 additional output lines can be written before returning. There is also an
8425 smtp_fflush() function to enable the detection of a dropped connection.
8426 The variables smtp_input and smtp_batched_input are exported to
8429 55. Changed the default runtime configuration: the message "Unknown user"
8430 has been removed from the ACL, and instead placed on the localuser router,
8431 using the cannot_route_message feature. This means that any verification
8432 failures that generate their own messages won't get overridden. Similarly,
8433 the "Unrouteable address" message that was in the ACL for unverifiable
8434 relay addresses has also been removed.
8436 56. Added hosts_avoid_esmtp to the smtp transport.
8438 57. The exicyclog script was not checking for the esoteric option
8439 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8440 will work only if exicyclog is run under the appropriate euid.
8442 58. Following a discussion on the list, the rules by which Exim recognises line
8443 endings on incoming messages have been changed. The -dropcr and drop_cr
8444 options are now no-ops, retained only for backwards compatibility. The
8445 following line terminators are recognized: LF CRLF CR. However, special
8446 processing applies to CR:
8448 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8449 nor a local message in the state where . is a terminator.
8451 (ii) If a bare CR is encountered in a header line, an extra space is added
8452 after the line terminator so as not to end the header. The reasoning
8453 behind this is that bare CRs in header lines are most likely either
8454 to be mistakes, or people trying to play silly games.
8456 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8457 was being incorrectly given as 18 bytes larger than it should have been.
8458 This is a VOB (very old bug).
8460 60. This may never have affected anything current, but just in case it has:
8461 When the local host is found other than at the start of a list of hosts,
8462 the local host, those with the same MX, and any that follow, are discarded.
8463 When the list in question was part of a longer list of hosts, the following
8464 hosts (not currently being processed) were also being discarded. This no
8465 longer happens. I'm not sure if this situation could ever has previously
8468 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8471 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8472 header. This is something that is recommended in a new Internet Draft, and
8473 is something that is documented as being done by Sendmail. There are two
8474 possible values. For messages generated by the autoreply transport, Exim
8477 Auto-Submitted: auto-replied
8479 whereas for all other generated messages (e.g. bounces) it adds
8481 Auto-Submitted: auto-generated
8483 63. The "personal" condition in filters now includes a test for the
8484 Auto-Submitted: header. If it contains the string "auto-" the message it
8485 not considered personal.
8487 64. Added rcpt_include_affixes as a generic transport option.
8489 65. Added queue_only_override (default true).
8491 66. Added the syslog_duplication option.
8493 67. If what should have been the first header line of a message consisted of
8494 a space followed by a colon, Exim was mis-interpreting it as a header line.
8495 It isn't of course - it is syntactically invalid and should therefore be
8496 treated as the start of the message body. The misbehaviour could have
8497 caused a number of strange effects, including loss of data in subsequent
8498 header lines, and spool format errors.
8500 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8501 client host had authenticated. This control can now be exercised by an ACL
8502 for more flexibility.
8504 69. By default, callouts do not happen when testing with -bh. There is now a
8505 variant, -bhc, which does actually run the callout code, including
8506 consulting and updating the callout cache.
8508 70. Added support for saslauthd authentication, courtesy of Alexander
8511 71. If statvfs() failed on the spool or log directories while checking their
8512 size for availability, Exim confusingly gave the error "space shortage".
8513 Furthermore, in debugging mode it crashed with a floating point exception.
8514 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8515 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8516 serious problem, Exim now writes to the main and panic logs when this
8517 happens, with details of the failure. It then refuses to accept the
8518 incoming message, giving the message "spool directory problem" or "log
8519 directory problem" with a 421 code for SMTP messages.
8521 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8522 0, 1, and 2 exist, because some OS complain for execs without them (see
8523 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8524 descriptors. However, the code omitted to check that the open succeeded,
8525 causing mysterious errors if for some reason the permissions on /dev/null
8526 got screwed. Now Exim writes a message to the main and panic logs, and
8527 bombs out if it can't open /dev/null.
8529 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8530 interact so that it is all more flexible. It is supposed to remain
8531 backwards compatible. Also added extra_local_interfaces.
8533 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8534 to bomb out with an assertion failure - to the client this appears as a
8535 connection drop. This problem occurs in the part of the code that was taken
8536 from the Samba project. Fortunately, the assertion is in a very simple
8537 function, so I have fixed this by reproducing the function inline in the
8538 one place where it is called, and arranging for authentication to fail
8539 instead of killing the process with assert().
8541 75. The SPA client code was not working when the server requested OEM rather
8542 than Unicode encoding.
8544 76. Added code to make require_files with a specific uid setting more usable in
8545 the case where statting the file as root fails - usually a non-root-mounted
8546 NFS file system. When this happens and the failure is EACCES, Exim now
8547 forks a subprocess and does the per-uid checking as the relevant uid.
8549 77. Added process_log_path.
8551 78. If log_file_path was not explicitly set, a setting of check_log_space or
8552 check_log_inodes was ignored.
8554 79. If a space check for the spool or log partitions fails, the incident is now
8555 logged. Of course, in the latter case the data may get lost...
8557 80. Added the %p formatting code to string_format() so that it can be used to
8558 print addresses in debug_print(). Adjusted all the address printing in the
8559 debugging in store.c to use %p rather than %d.
8561 81. There was a concern that a line of code in smtp_in.c could overflow a
8562 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8563 initially expressed, the concern was not well-founded, because trailing
8564 spaces are removed early. However, if the trailing spaces were followed by
8565 a NULL, they did not get removed, so the overflow was possible. Two fixes
8568 (a) I re-wrote the offending code in a cleaner fashion.
8569 (b) If an incoming SMTP command contains a NULL character, it is rejected
8572 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8573 runs initgroups(), so that if the Exim user is in any additional groups,
8574 they will be used during message reception.
8580 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob