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.
96 JH/20 Bug 3047: A recent (somewhere between 10.34 and 10.42) version of the
97 pcre2 library starting allocating 20kB rather than 112 bytes per match
98 call, which broke the 2GB total limitation on Exim's memory management
99 when a user had over 104207 messages stored and the appendfile
100 maildir_quota_directory_regex option is in use. Release the allocated
101 memory every thosand files to avoid this.
102 The same issue arises with the ACL regex condition, which is applied
103 to every line of a received message.
109 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
110 SMTP connection" log lines.
112 JH/02 Option default value updates:
113 - queue_fast_ramp (main) true (was false)
114 - remote_max_parallel (main) 4 (was 2)
116 JH/03 Cache static regex pattern compilations, for use by ACLs.
118 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
119 Make the rewrite never match and keep the logging. Trust the
120 admin to be using verify=header-syntax (to actually reject the message).
122 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
123 (under Linux) we watch the dir containing the final file; previously
124 it would be the dir with the first symlink. We still do not monitor
127 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
128 dn_expand() is happy to pass them through.
130 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
131 the file with the proof had an unchanged name, the new proof(s) were
132 loaded on top of the old ones (and nover used; the old ones were stapled).
134 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
135 more than one message arrived in a single connection a reference from
136 the earlier message could be re-used. Often a sigsegv resulted.
137 These variables were introduced in Exim 4.87.
138 Debug help from Graeme Fowler.
140 JH/09 Fix ${filter } for conditions that modify $value. Previously the
141 modified version would be used in construction the result, and a memory
144 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
145 Find and fix by Jasen Betts.
147 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
148 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
149 the systemwide configuration to override the Exim config.
151 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
152 API changes in libopendmarc.
154 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
155 pid 1, in the normal "background daemon" mode, having to drop process-
156 group leadership also lost track of needing to create listener sockets.
158 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
159 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
161 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
162 a capture group which obtained no text (eg. "(abc)*" matching zero
163 occurrences) could cause a segfault if the corresponding $<n> was
166 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
167 included a close-brace character (eg. it itself used an expansion) an
170 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
171 starting TLS. Previously it was after, meaning that attackers on such
172 ports had to be screened using the host_reject_connection main config
173 option. The new sequence aligns better with the STARTTLS behaviour, and
174 permits defences against crypto-processing load attacks, even though it
175 is strictly an incompatible change.
176 Also, avoid sending any SMTP fail response for either the connect ACL
177 or host_reject_connection, for TLS-on-connect ports.
179 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
180 Previously this was not permitted, but it makes reasonable sense.
181 While there, restore a restriction on using it from a connect ACL; given
182 the change JH/16 it could only return false (and before 4.91 was not
185 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
186 was exactly sized compared to the log buffer, a crash occurred with the
187 misleading message "bad memory reference; pool not found".
188 Found and traced by Jasen Betts.
190 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
191 dns_again_means_nonexist included an element causing a DNS lookup which
192 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
193 included (though probably not limited to) a process crash from stack
194 memory limit, or from excessive open files. Replace this with a paniclog
195 whine (as this is likely a configuration error), and returning
198 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
199 this always failed, probably leading to the usual downgrade to in-clear
202 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
203 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
204 to not checking that list for these lookups.
206 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
209 JH/23 Fix crash in string expansions. Previously, if an empty variable was
210 immediately followed by an expansion operator, a null-indirection read
211 was done, killing the process.
213 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
214 include an SMTP response string which is longer than that supported
215 by the delivering transport. Alleviate by wrapping such lines before
218 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
219 chars (RFC limit). Previously a limit of 12 items was made, which with
220 a not-impossible References: in the message being bounced could still
221 be over-large and get stopped in the transport.
223 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
224 close. Previously a bare socket close was done.
226 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
229 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
230 message being considered for delivery was already being handled by
231 another process, and having an SMTP connection already open, the function
232 to close it tried to use an uninitialized variable. This would afftect
233 high-volume sites more, especially when running mailing-list-style loads.
234 Pollution of logs was the major effect, as the other process delivered
235 the message. Found and partly investigated by Graeme Fowler.
237 JH/29 Change format of the internal ID used for message identification. The old
238 version only supported 31 bits for a PID element; the new 64 (on systems
239 which can use Base-62 encoding, which is all currently supported ones
240 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
241 and must use Base-36). The new ID is 23 characters rather than 16, and is
242 visible in various places - notably logs, message headers, and spool file
243 names. Various of the ancillary utilities also have to know the format.
244 As well as the expanded PID portion, the sub-second part of the time
245 recorded in the ID is expanded to support finer precision. Theoretically
246 this permits a receive rate from a single comms channel of better than the
248 The major timestamp part of the ID is not changed; at 6 characters it is
249 usable until about year 3700.
250 Updating from previously releases is fully supported: old-format spool
251 files are still usable, and the utilities support both formats. New
252 message will use the new format. The one hints-DB file type which uses
253 message-IDs (the transport wait- DB) will be discarded if an old-format ID
254 is seen; new ones will be built with only new-format IDs.
255 Optionally, a utility can be used to convert spool files from old to new,
256 but this is only an efficiency measure not a requirement for operation
257 Downgrading from new to old requires running a provided utility, having
258 first stopped all operations. This will convert any spool files from new
259 back to old (losing time-precision and PID information) and remove any
260 wait- hints databases.
262 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
263 we treated them as item separators when parsing for a list item, but they
264 need to be protected by the doublequotes. While there, add handling for
267 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
268 Found and fixed by Jasen Betts. No testcase for this as my usual text
269 editor insists on emitting only valid UTF-8.
271 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
272 a null-indirection SIGSEGV for the receive process.
274 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
275 -bh use. Internal checking would spot this and take a panic.
277 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
278 In 4.96 this would expand to empty.
280 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
281 certificate. Find and fix by Andreas Metzler.
283 JH/36 Add ARC info to DMARC hostory records.
285 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
286 or fakedefer. Previously the sender could discover that the message
287 had in fact been accepted.
289 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
290 sequences. Previously the input was not noted as being tainted; notably
291 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
292 bad coding of authenticators.
294 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
295 and ${tr...}. Found and diagnosed by Heiko Schlichting.
297 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
298 could be triggered by externally-supplied input. Found by Trend Micro.
301 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
302 be triggered by externally-controlled input. Found by Trend Micro.
305 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
306 be triggered by externally-controlled input. Found by Trend Micro.
309 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
310 Make the rewrite never match and keep the logging. Trust the
311 admin to be using verify=header-syntax (to actually reject the message).
313 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
316 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
322 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
323 after reception to before a subsequent reception. This should
324 mean slightly faster delivery, and also confirmation of reception
327 JH/02 Move from using the pcre library to pcre2. The former is no longer
328 being developed or supported (by the original developer).
330 JH/03 Constification work in the filters module required a major version
331 bump for the local-scan API. Specifically, the "headers_charset"
332 global which is visible via the API is now const and may therefore
333 not be modified by local-scan code.
335 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
336 sendfile() didi not account for the way the ClamAV driver code called it.
338 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
339 time check was being done for every character; replace that with one
342 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
343 sent was prefixed with a length byte.
345 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
346 RFC 5321. Previously Dovecot (at least) would log errors during
349 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
350 in 4.61 without a reason logged. Binaries will be bigger, which might
351 matter on diskspace-constrained systems, but debug is easier.
353 JH/09 Fix macro-definition during "-be" expansion testing. The move to
354 write-protected store for macros had not accounted for these runtime
355 additions; fix by removing this protection for "-be" mode.
357 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
358 handing out large-numbered file descriptors, violating the usual Unix
359 assumption (and required by Posix) that the lowest possible number will be
360 allocated by the kernel when a new one is needed. In the daemon, and any
361 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
362 are not useable for FD_SET() [and hence select()] and overwrite the stack.
363 Assorted crashes happen.
365 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
366 main-section options or in a connect ACL, the value from the first ever
367 connection was never replaced for subsequent connections. Found by
370 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
371 though only once PCRE2 was introduced: the memory accounting used under
372 debug offset allocations by an int, giving a hard trap in early startup.
373 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
375 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
376 with underbars is given. The write-protection of configuration introduced
377 in 4.95 trapped when normalisation was applied to an option not needing
380 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
382 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
383 failing to close the certificates directory, every hour or any time it
386 JH/16 Debugging initiated by an ACL control now continues through into routing
387 and transport processes. Previously debugging stopped any time Exim
388 re-execs, or for processing a queued message.
390 JH/17 The "expand" debug selector now gives more detail, specifically on the
391 result of expansion operators and items.
393 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
394 bad comparison between the option value and the name of the file to
395 be included was done, and a mismatch was wrongly identified.
396 4.88 to 4.95 are affected.
398 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
400 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
401 supplied as the db file-pair basename. Previously, if a directory
402 path was given, for example via the autoreply "once" option, the DB
403 file.pag and file.dir files would be created in that directory's
406 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
407 "taint" log_selector. These were previously deprecated.
409 JH/22 Fix static address-list lookups to properly return the matched item.
410 Previously only the domain part was returned.
412 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
413 the call into OpenSSL to send a TLS Close was being repeated; this
414 resulted in the library waiting for the peer's Close. If that was never
415 sent we waited forever. Fix by tracking send calls.
417 JH/24 The ${run} expansion item now expands its command string elements after
418 splitting. Previously it was before; the new ordering makes handling
419 zero-length arguments simpler. The old ordering can be obtained by
420 appending a new option "preexpand", after a comma, to the "run".
422 JH/25 Taint-check exec arguments for transport-initiated external processes.
423 Previously, tainted values could be used. This affects "pipe", "lmtp" and
424 "queryprogram" transport, transport-filter, and ETRN commands.
425 The ${run} expansion is also affected: in "preexpand" mode no part of
426 the command line may be tainted, in default mode the executable name
429 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
430 the the facility was not passed across execs, and only the first message
431 passed over a connection could use BDAT; any further ones using DATA.
433 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
434 uses $sending_ip_address and an interface is specified.
435 Previously any use of the local address in the EHLO name disabled
436 PIPECONNECT, the common case being to use the rDNS of it.
438 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
439 resumption. Previously verify failed because no certificate status is
440 passed on the wire for the restarted session. Fix by using the recorded
441 ocsp status of the stored session for the new connection.
443 JH/29 TLS resumption: the key for session lookup in the client now includes
444 more info that a server could potentially use in configuring a TLS
445 session, avoiding oferring mismatching sessions to such a server.
446 Previously only the server IP was used.
448 JH/30 Fix string_copyn() for limit greater than actual string length.
449 Previously the copied amount was the limit, which could result in a
450 overlapping memcpy for newly allocated destination soon after a
451 source string shorter than the limit. Found/investigated by KM.
453 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
454 close; it may be needed for a subsequent connection. This caused a
455 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
457 JH/32 Fix CHUNKING for a second message on a connection when the first was
458 rejected. Previously we did not reset the chunking-offered state, and
459 erroneously rejected the BDAT command. Investigation help from
462 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
463 an empty address. Previously the expansion returned an error.
465 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
466 proxy. Previously these were misparsed, leading to paniclog entries.
472 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
473 related applications. Previously an "H" was used where available info
474 says that "M" should be, so change to match.
476 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
477 as arguments, so an implementation trying to copy these into a local
478 buffer was taking a taint-enforcement trap. Fix by using dynamically
479 created buffers. Similar fix for radius expansion condition.
481 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
482 reasonable, eg. to count headers. Fix by using dynamically created
483 buffers rather than a local. Do similar fixes for ACL actions "dcc",
484 "log_reject_target", "malware" and "spam"; the arguments are expanded
485 so could be handling tainted values.
487 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
488 broken the (no-op) support for this sendmail command. Restore it
489 to doing nothing, silently, and returning good status.
491 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
492 record path was given (or the default used) without a leading directory
493 path, an error occurred on trying to open it. Use the transport's working
496 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
497 transport. Previously it was the name on the DNS A-record; use instead
498 the head of the CNAME chain leading there (if there is one). This seems
499 to align better with RFC 6125.
501 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
502 smtp_accept_max_per_host allocated resources which were not released
503 when the limit was exceeded. This eventually crashed the daemon. Fix
504 by adding a release action in that path.
506 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
507 expanded; previously using tainted values was rejected. Fix by using
508 dynamically-created buffers.
510 JH/09 Relax restrictions on ACL verify condition needing access to message
511 headers. Previously they were only permitted in data and non-smtp ACLs;
512 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
513 not_blind, header_sender and header_names_ascii verification.
515 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
516 Previously a macro used one argument twice; when called with the
517 argument as an expression having side-effects, incorrect operation
518 resulted. Use an inlineable function.
520 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
521 held open for a verify callout. Previously this wan not accounted for
522 and a corrupt onward SMTP conversation resulted.
524 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
525 Previously they were embedded naively in the constructed address; when
526 needed, strip the quoting and quote the entire local_part.
527 Also make the inbound_srs expansion condition handle quoting.
529 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
530 excluded, not matching the documentation.
532 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
533 was given for the sqlite_dbfile a trap resulted.
535 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
536 "name" argument resulted in a trap. There is no reason to disallow such;
537 this was a coding error.
539 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
540 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
541 spent suspended, ignoring the POSIX definition. Previously we assumed
542 it did and a constant offset from real time could be used as a correction.
543 Change to using the same clock source for the start-of-message and the
544 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
545 exists, just to get a clock slightly more aligned to reality.
547 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
548 RFC says it is optional some validators care. The missing char was not
549 intended but triggered by a line-wrap alignment. Discovery and fix by
550 Guillaume Outters, hacked on by JH.
552 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
553 name being quoted was tainted a trap would be taken. Fix by using
554 dynamically created buffers. The routine could have been called by a
555 rewrite with the "h" flag, by using the "-F" command-line option, or
556 by using a "name=" option on a control=submission ACL modifier.
558 JH/19 SPF: change the Authentication-Results expansion component to give
559 smtp.helo when the sender domain is empty. Previously it gave
562 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
563 not in 127.0.0.0/8 to help in spotting list domains taken over by a
564 domain-parking registrar.
566 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
567 Previously when a whitespace character was specified it was not inserted
568 after removing the newline.
570 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
571 the domain part of the recipient address. This overrides any tls_sni
572 option set, which was previously used.
574 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
577 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
578 is_tainted() had an off-by-one error in the overenthusiastic direction.
579 Find and fix by Gavan. Although NetBSD is not a supported platform for
580 4.94 this bug could affect other platforms.
582 PP/01 Fix default prime selection to be consistent.
583 One path used ike23 still, instead of exim.dev.20160529.3; now both
584 execution flows will use the same DH primes (currently
585 exim.dev.20160529.3).
587 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
588 option in smtp transport, to match the documentation. Previously
589 verification was not being done in some cases where it should have been.
591 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
592 than one server was defined and depending on the platform memory layout
593 details, an internal consistency trap could be hit while walking the list
596 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
597 files. The value is used by the authresults expansion item. Previously
598 if this was used in a router or transport, a crash could result.
600 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
601 have a sufficient incidence of EINTR returns from select that an
602 interaction with other operations done by the main daemon loop exposed
603 a bug in the error-handling. This was benign apart from the log
606 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
607 for consistency with delivered "=>" and failed "**" lines. While we're
608 there, handle PRX and TFO.
610 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
611 applied. This resulted, if any header-line rewrite rules were configured,
612 in a panic-log triggerable by sending a message with a long address in
613 a header. Fix by increasing the arbitrary limit to larger than a single
614 (dewrapped) 5322 header line maximum size.
616 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
617 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
618 RFC 6648 which deprecates X- options in protocols as a general practice.
619 Changeover between the implementations is handled by the mechanisms
622 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
623 a remote which uses callout/hold. Previously the local was queued.
625 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
628 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
629 log line. The asterisk is now consistently appended to the remote IP
630 (and port, if given), and will also be provided on defer and fail log
631 lines. Previously it could be placed on the local IP if that was being
632 logged, and was only provided on delivery lines.
634 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
636 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
637 authenticator client_send option. Previously the next char, after a pair
638 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
639 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
640 documentation. There is still no way to get a leading ^ immediately
641 after a NUL (ie. for the password of a PLAIN method authenticator.
643 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
644 files. For bad sizes read, delete the record and whine to paniclog.
646 JH/38 When logging an AUTH failure, as server, do not include sensitive
647 information. Previously, the credentials would be included if given
648 as part of the AUTH command line and an ACL denied authentication.
650 JH/39 Bug 2691: fix $local_part_data. When the matching list element
651 referred to a file, bad data was returned. This likely also affected
654 JH/40 The gsasl authenticator now supports caching of the salted password
655 generated by the client-side implementation. This required the addition
656 of a new variable: $auth4.
658 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
659 left undeleted; the attempt to re-create it then failed - resulting in
660 the usual "SIGHUP tp have daemon reload configuration" to not work.
661 This affected any platform not supporting "abstract" Unix-domain
662 sockets (i.e. not Linux).
664 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
665 recipients" response to RCPT in a later response, with a 250. The
666 previous coding assumed this would not happen, and under PIPELINING
667 would result in both lost and duplicate recipients for a message.
669 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
670 Previously the weighting was incorrectly applied. Similar fix for socks
671 proxies. Found and fixed by Heiko Schlichting.
673 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
674 not handle sub-lists included using the +namedlist syntax. While
675 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
676 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
679 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
680 Previously the permanent pool was used, so the sore could not be freed.
681 This meant a connection with many messages would use continually-growing
684 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
685 per-pool so as not to waste too much space. Previously a constant size
686 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
687 DOS attacks harder. The cost is wasted memory use in the larger blocks.
689 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
690 same space repeatedly, and a smaller process footprint.
692 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
693 for GnuTLS. Previously the name started "%s" which, while not a bug,
694 looked as if if might be one.
696 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
697 "continued connection") make the $proxy_* variables available. Previously
698 the information was not passed across the exec() call for subsequent
699 transport executions. This also mean that the log lines for the
700 messages can show the proxy information.
702 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
703 receive time. With modern systems the difference is significant.
704 The historical behaviour can be restored by disabling (a new) log_selector
705 "queue_time_exclusive".
707 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
708 content) were specifically tested for. Now, also cover expansions
709 resulting in ACL names and inline ACL content.
711 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
712 making it unusable in complex expressions.
714 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
715 fail with a taint-check on the spoolfile name, and leave the message
718 HS/01 Enforce absolute PID file path name.
720 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
722 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
724 PP/02 Bug 2643: Correct TLS DH constants.
725 A missing NUL termination in our code-generation tool had led to some
726 incorrect Diffie-Hellman constants in the Exim source.
727 Reported by kylon94, code-gen tool fix by Simon Arlott.
729 PP/03 Impose security length checks on various command-line options.
730 Fixes CVE-2020-SPRSS reported by Qualys.
732 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
733 better. Reported by Qualys.
735 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
736 providing a particularly obnoxious sender full name.
739 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
741 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
744 PP/08 Change default for recipients_max from unlimited to 50,000.
746 PP/09 Fix security issue with too many recipients on a message (to remove a
747 known security problem if someone does set recipients_max to unlimited,
748 or if local additions add to the recipient list).
749 Fixes CVE-2020-RCPTL reported by Qualys.
751 PP/10 Fix security issue in SMTP verb option parsing
752 Fixes CVE-2020-EXOPT reported by Qualys.
754 PP/11 Fix security issue in BDAT state confusion.
755 Ensure we reset known-good where we know we need to not be reading BDAT
756 data, as a general case fix, and move the places where we switch to BDAT
757 mode until after various protocol state checks.
758 Fixes CVE-2020-BDATA reported by Qualys.
760 HS/03 Die on "/../" in msglog file names
762 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
763 the Exim runtime user are allowed to create files.
765 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
768 QS/03 When reading the output from interpreted forward files we do not
769 pass the pipe between the parent and the interpreting process to
770 executed child processes (if any).
772 QS/04 Always die if requested from internal logging, even is logging is
775 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
776 the API; compilation noo longer completes with DMARC support included.
777 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
778 been reported on other platforms.
780 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
782 JH/56 Make the majority of info read from config files readonly, for defence-in-
783 depth against exploits. Suggestion by Qualys.
784 Not supported on Solaris 10.
786 JH/57 Fix control=fakreject for a custom message containing tainted data.
787 Previously this resulted in a log complaint, due to a re-expansion present
788 since fakereject was originally introduced.
790 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
791 modifier was given, a loop resulted.
793 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
794 pipelined SMTP command followed the BDAT LAST then it would be
795 incorrectly treated as part of the message body, causing a verification
798 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
799 require looser validation rules than those for 821-level addresses,
800 which only permit IP addresses.
806 JH/01 Avoid costly startup code when not strictly needed. This reduces time
807 for some exim process initialisations. It does mean that the logging
808 of TLS configuration problems is only done for the daemon startup.
810 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
812 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
813 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
816 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
817 DKIM signing is being done. Previously a transport_filter would always
818 disable CHUNKING, falling back to traditional DATA.
820 JH/05 Regard command-line recipients as tainted.
822 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
824 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
825 PAM library frees one of the arguments given to it, despite the
826 documentation. Therefore a plain malloc must be used.
828 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
829 on-stack buffers were used, resulting in a taint trap when DSN information
830 copied from a received message was written into the buffer.
832 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
833 the ordering of its ARC headers. This caused a crash.
835 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
836 a new record was being constructed with information from the peer, a trap
839 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
840 installation would get error messages from DMARC verify, when it hit the
841 nonexistent file indicated by the default. Distros wanting DMARC enabled
842 should both provide the file and set the option.
843 Also enforce no DMARC verification for command-line sourced messages.
845 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
846 could, depending on the platform, hang at the STARTTLS response.
848 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
849 message on a connection. Previously if one message had ARC headers and
850 the following one did not, a crash could result when adding an
851 Authentication-Results: header.
853 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
854 Exim main code and Exim-related utities. The introduction of taint
855 tracking also did many adjustments to string handling. Since then, eximon
856 frequently terminated with an assert failure.
858 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
859 check for 452 responses. This slightly helps the inefficieny of doing
860 a large alias-expansion into a recipient-limited target. The max_rcpt
861 transport option still applies (and at the current default, will override
862 the new feature). The check is done for either cause of synch, and forces
863 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
864 connection. The new facility is not tunable at this time.
866 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
867 library live data was being used, so the results became garbage. Make
868 copies while it is still usable.
870 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
871 on delivery deferred (==) and failed (**) lines (if a delivery was
872 attemtped). Previously it was only on completion (=>) lines.
874 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
875 for the expansion of the server_scram_iter and server_scram_salt options.
877 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
878 are now specifically given a NO_DATA response without hitting the system
879 resolver. The library goes on to do the now-standard TXT lookup.
880 Use of dnsdb lookups is not affected.
882 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
883 only retrieve the errormessage once. Previously two calls to dlerror()
884 were used, and the second one (for mainlog/paniclog) retrieved null
887 JH/20 Taint checking: disallow use of tainted data for
888 - the appendfile transport file and directory options
889 - the pipe transport command
890 - the autoreply transport file, log and once options
891 - file names used by the redirect router (including filter files)
893 - paths used by single-key lookups
894 Previously this was permitted.
896 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
897 adjusted the size of a major service buffer; this failed because the
898 buffer was in use at the time. Change to a compile-time increase in the
899 buffer size, when this authenticator is compiled into exim.
901 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
902 previous fast-mode was untenable in the face of glibs using mmap to
903 support larger malloc requests.
905 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
906 New values supported, if defined on system where compiled:
907 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
908 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
910 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
911 running a limited number of proceses in parallel, a benefit is gained. The
912 amount varies with the platform hardware and load. The use of the option
913 queue_run_in_order means we cannot do this, as ordering becomes
916 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
917 had introduced a string-copy (for ensuring NUL-termination) which was not
918 appropriate for that case, which can include embedded NUL bytes in the
919 block of data. Investigation showed the copy to actually be needless, the
920 data being length-specified.
922 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
923 done during a receiving connection, and both used TLS, global info was
924 used rather than per-connection info for tracking the state of data
925 queued for transmission. This could result in a connection hang.
927 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
928 Previously, when delivering serveral messages down a single connection
929 only the first would provide a SIZE. This was due to the size information
930 not being properly tracked.
932 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
933 TAI (at 37 seconds currently), pretend to be in UTC for time-related
934 expansion and logging. Previously, spurious values such as a future
935 minute could be seen.
937 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
938 it could crash from a null-deref. This could also affect the
939 ${addresses: } operator and ${readsock } item.
941 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
942 message following a mime one, the variable was not reset.
944 JH/30 When an pipelined-connect fails at the first response, assume incorrect
945 cached capability (perhaps the peer reneged?) and immediately retry in
948 JH/31 Fix spurious detection of timeout while writing to transport filter.
950 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
951 an attempt to copy the string was made before checking it.
953 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
954 taint of the lookup key was maintained; we now regard the presence in the
955 filesystem as sufficient validation.
957 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
958 argument is supplied.
960 JH/35 The dsearch lookup now requires that the directory is an absolute path.
961 Previously this was not checked, and nonempty relative paths made an
962 access under Exim's current working directory.
964 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
965 Previously no event was raised.
967 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
968 parameter supplied by the sender MAIL FROM command. Previously it was
969 ignored, and only the check_spool_space option value for the required
972 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
973 the size of the signing public-key. Previously it was instead giving
974 the size of the signature hash.
976 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
977 the default. See the (new) dkim_verify_min_keysizes option.
979 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
980 and an ACL use a lookup for checking either the local_part or domain,
981 stale data could be accessed. Ensure that variable references are
982 dropped between messages.
984 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
985 by the client was not checked as pointing within response data before
986 being used. A malicious client could thus cause an out-of-bounds read and
987 possibly gain authentication. Fix by adding the check.
989 JH/42 Internationalisation: change the default for downconversion in the smtp
990 transport to be "if needed". Previously it was "as previously set" for
991 the message, which usually meant "if needed" for message-submission but
992 "no" for everything else. However, MTAs have been seen using SMTPUTF8
993 even when the envelope addresses did not need it, resulting in forwarding
994 failures to non-supporting MTAs. A downconvert in such cases will be
995 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
996 transport. The change does mean that addresses needing conversion will
997 be converted when previously a delivery failure would occur.
999 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
1000 response was received it would be used unchecked in a fail-DSN, violating
1001 standards on line-length limits. Truncate if needed.
1003 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1004 doesn't work. (Additionally add a new main config option to configure the
1011 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1012 side, to decode a TLS 1.3 packet capture.
1014 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1015 Previously the default library behaviour applied, sending two, each in
1016 its own TCP segment.
1018 JH/03 Debug output for ACL now gives the config file name and line number for
1021 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1023 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1025 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1026 buffer overrun for (non-chunking) other transports.
1028 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1029 TLS1.3, means that a server rejecting a client certificate is not visible
1030 to the client until the first read of encrypted data (typically the
1031 response to EHLO). Add detection for that case and treat it as a failed
1032 TLS connection attempt, so that the normal retry-in-clear can work (if
1033 suitably configured).
1035 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1036 and/or domain. Found and fixed by Jason Betts.
1038 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1039 configuration). If a CNAME target was not a wellformed name pattern, a
1042 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1043 the OS reports them interleaved with other addresses.
1045 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1046 used both for input and for a verify callout, both encrypted, SMTP
1047 responses being sent by the server could be lost. This resulted in
1048 dropped connections and sometimes bounces generated by a peer sending
1051 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1052 string. Previously it was possible to cause undefined behaviour in a
1053 library routine (usually a crash). Found by "zerons".
1055 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1058 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1059 API was removed, so update to use the newer ones.
1061 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1062 any timeout set, is taking a long time. Previously we would hang on to a
1063 rotated logfile "forever" if the input was arriving with long gaps
1064 (a previous attempt to fix addressed lack, for a long time, of initial
1067 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1068 shared (NFS) environment. The length of the tempfile name is now
1069 4 + 16 ("hdr.$message_exim_id") which might break on file
1070 systems which restrict the file name length to lower values.
1071 (It was "hdr.$pid".)
1073 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1074 shared (NFS) environment.
1076 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1077 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1080 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1081 on some platforms for bit 31.
1083 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1084 to changes apparently associated with TLS1.3 handling some of the APIs
1085 previously used were either nonfunctional or inappropriate. Strings
1086 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1087 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1088 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1089 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1090 and the use of specific cipher names in the encrypted= ACL condition.
1092 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1094 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1095 verification result was not updated unless hosts_require_ocsp applied.
1097 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1098 queue_list_requires_admin set to false, non-admin users were denied the
1101 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1102 directory-of-certs mode. Previously they were advertised despite the
1105 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1106 A single TCP connection by a client will now hold a TLS connection open
1107 for multiple message deliveries, by default. Previously the default was to
1110 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1111 default. If built with the facility, DANE will be used. The facility
1112 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1114 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1115 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1116 must be defined and you must still, unless you define DISABLE_TLS, manage
1117 the the include-dir and library-file requirements that go with that
1118 choice. Non-TLS builds are still supported.
1120 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1123 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1124 default. If the platform supports and has the facility enabled, it will
1125 be requested on all coneections.
1127 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1128 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1130 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1132 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1133 success-DSN messages. Previously the From: header was always the default
1134 one for these; the option was ignored.
1136 JH/28 Fix the timeout on smtp response to apply to the whole response.
1137 Previously it was reset for every read, so a teergrubing peer sending
1138 single bytes within the time limit could extend the connection for a
1139 long time. Credit to Qualsys Security Advisory Team for the discovery.
1141 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1142 delivery address, which leaked information of the results of local
1143 forwarding. Change to the original envelope recipient address, per
1146 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1147 requested. Previously not bounce was generated and a log entry of
1148 error ignored was made.
1150 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1152 JH/32 Introduce a general tainting mechanism for values read from the input
1153 channel, and values derived from them. Refuse to expand any tainted
1154 values, to catch one form of exploit.
1156 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1157 was unused and the unexpanded text used for the test. Found and
1158 fixed by Ruben Jenster.
1160 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1161 an attempt to use a TLS library read routine dereffed a nul pointer,
1164 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1165 them in our smtp response.
1167 JH/36 Have the general router option retry_use_local_part default to true when
1168 any of the restrictive preconditions are set (to anything). Previously it
1169 was only for check_local user. The change removes one item of manual
1170 configuration which is required for proper retries when a remote router
1171 handles a subset of addresses for a domain.
1173 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1174 link count into consideration.
1176 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1177 caused the extension of big_buffer, the following lines were ignored.
1179 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1180 accordance with RFC 2308. Previously there was no expiry, so a longlived
1181 receive process (eg. due to ACL delays) versus a short SOA value could
1184 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1186 JH/39 Promote DMARC support to mainline.
1188 JH/40 Bug 2452: Add a References: header to DSNs.
1190 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1191 parameters. The relevant library call is documented as "Deprecated: This
1192 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1193 3.6.0, DH parameters are negotiated following RFC7919."
1195 HS/06 Change the default of dnssec_request_domains to "*"
1197 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1198 carried on and emitted a BDAT command, even when PIPELINING was not
1201 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1202 buffer was used for the filename, resulting in a trap when tainted
1203 arguments (eg. $domain) were used.
1205 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1206 recommended to avoid a possible server-load attack. The feature can be
1207 re-enabled via the openssl_options main cofiguration option.
1209 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1210 for version 4.90 - adding a "more data" boolean to the arguments.
1211 Bumped the ABI version number also, this having been missed previously;
1212 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1213 respect of usage of smtp_printf() by either local_scan code or libraries
1214 accessed via the ${dlfunc } expansion item. Both will need coding
1215 adjustment for any calls to smtp_printf() to match the new function
1216 signature; a FALSE value for the new argument is always safe.
1218 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1219 the file-offset (which the Linux syscall does, and exim expects); this
1220 resulted in an indefinite loop.
1222 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1223 to do ARC verification. The Authentication-Results: header line added
1224 by the configuration then had no ARC item.
1226 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1227 an incorrect mode was used for reading the result, resulting in it being
1234 JH/01 Remove code calling the customisable local_scan function, unless a new
1235 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1237 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1238 non-signal-safe functions being used.
1240 JH/03 Bug 2269: When presented with a received message having a stupidly large
1241 number of DKIM-Signature headers, disable DKIM verification to avoid
1242 a resource-consumption attack. The limit is set at twenty.
1244 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1245 report of oldest_pass in ${authres } in consequence, and separate out
1246 some descriptions of reasons for verification fail.
1248 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1249 files in the spool were present and unlocked. A queue-runner could spot
1250 them, resulting in a duplicate delivery. Fix that by doing the unlock
1251 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1252 add more error-checking on spoolfile handling while that code is being
1255 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1256 No known attacks, no CVE, this is defensive hardening.
1258 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1259 a queue-runner could start a delivery while other operations were ongoing.
1260 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1261 Found and investigated by Tim Stewart. Fix by using the open message data
1262 file handle rather than opening another, and not locally closing it (which
1263 releases a lock) for that case, while creating the temporary .eml format
1264 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1266 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1267 $sender_verify_failure/$recipient_verify_failure to "random".
1269 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1272 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1273 Previously this would segfault.
1275 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1278 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1279 like zero, since the resolver should be doing this for us, But we need one
1280 as a CNAME but no MX presence gets the CNAME returned; we need to check
1281 that doesn't point to an MX to declare it "no MX returned" rather than
1282 "error, loop". A new main option is added so the older capability of
1283 following some limited number of chain links is maintained.
1285 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1287 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1288 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1289 now-deprecated earlier definitions used only specified the range up to TLS
1290 1.2 (in the older-version library docs).
1292 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1294 JH/15 Rework TLS client-side context management. Stop using a global, and
1295 explicitly pass a context around. This enables future use of TLS for
1296 connections to service-daemons (eg. malware scanning) while a client smtp
1297 connection is using TLS; with cutthrough connections this is quite likely.
1299 JH/16 Fix ARC verification to do AS checks in reverse order.
1301 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1303 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1304 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1305 Previously the "utf8" would be re-prepended for every additional message.
1307 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1308 Previously thery were accepted, resulting in issues when attempting to
1309 forward messages to a non-supporting MTA.
1311 PP/02 Let -n work with printing macros too, not just options.
1313 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1314 one parent address was copied, and bogus data was used at delivery-logging
1315 time. Either a crash (after delivery) or bogus log data could result.
1316 Discovery and analysis by Tim Stewart.
1318 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1319 Previously if the string ended mid-character, we did not insert the
1320 promised '?' replacement.
1322 PP/04 Documentation: current string operators work on bytes, not codepoints.
1324 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1325 should pack well giving a smaller memory footprint so better caching and
1326 therefore performance. Group the declarations where this can't be done so
1327 that the byte-sized flag variables are not interspersed among pointer
1328 variables, giving a better chance of good packing by the compiler.
1330 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1331 non-null, to avoid issues with sites running BATV. Previously reports were
1332 sent with an empty envelope sender so looked like bounces.
1334 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1335 The ignore_error flag wasn't being returned from the filter subprocess so
1336 was not set for later routers. Investigation and fix by Matthias Kurz.
1338 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1339 and a msg:complete for the whole, when a message is manually removed using
1340 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1342 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1343 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1344 hope that is portable enough.
1346 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1347 requiring a known-CA anchor certificate; make it now rely entirely on the
1348 TLSA as an anchor. Checking the name on the leaf cert against the name
1349 on the A-record for the host is still done for TA (but not for EE mode).
1351 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1352 included in delivery lines for non-proxied connections, when compiled with
1353 SUPPORT_SOCKS and running with proxy logging enabled.
1355 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1356 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1357 move the existing event to fire before the normal logging of message
1358 failure so that custom logging is bracketed by normal logging.
1360 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1361 msg:fail:internal event. Developement by Matthias Kurz.
1363 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1364 far too small for todays use of crypto signatures stored there. Go all
1365 the way to the max DNS message size of 64kB, even though this might be
1366 overmuch for IOT constrained device use.
1368 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1369 copy a string over itself. The library routine is documented as not
1370 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1372 JH/32 For main options check_spool_space and check_inode_space, where the
1373 platform supports 64b integers, support more than the previous 2^31 kB
1374 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1375 the previous G, M, k.
1377 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1378 $authenticated_fail_id variable on authentication failure. Previously
1381 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1382 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1383 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1384 GNUTLS_SEC_PARAM_MEDIUM.
1386 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1387 side. Previously we would continue as if no SNI had been received.
1389 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1390 "<" character, which should be a list-separator specification, we walked
1391 off past the nul-terimation.
1393 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1394 causes) even when the retry time is not yet met. Previously they were
1395 not, meaning that when (say) an account was over-quota and temp-rejecting,
1396 and multiple senders' messages were queued, only one sender would get
1397 notified on each configured delay_warning cycle.
1399 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1401 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1402 cipher-suites, an error can be left on the stack even for a succeeding
1403 accept; this results in impossible error messages when a later operation
1406 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1407 return error codes indicating retry. Under TLS1.3 this becomes required.
1409 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1410 it only wrote the new authenticators, resulting in a lack of tracking of
1411 peer changes of ESMTP extensions until the next cache flush.
1413 JH/41 Fix the loop reading a message header line to check for integer overflow,
1414 and more-often against header_maxsize. Previously a crafted message could
1415 induce a crash of the recive process; now the message is cleanly rejected.
1417 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1418 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1424 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1425 When redis_servers is set to a list of > 1 element, and the Redis servers
1426 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1427 case of MOVED into a DEFER case instead, thus moving the query onto the
1428 next server in the list. For a cluster of N elements, all N servers must
1429 be defined in redis_servers.
1431 GF/02 Catch and remove uninitialized value warning in exiqsumm
1432 Check for existence of @ARGV before looking at $ARGV[0]
1434 JH/01 Replace the store_release() internal interface with store_newblock(),
1435 which internalises the check required to safely use the old one, plus
1436 the allocate and data copy operations duplicated in both (!) of the
1437 extant use locations.
1439 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1440 modifier. This matches the restriction on the commandline.
1442 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1443 Previously only the last row was returned.
1445 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1446 we assumed that tags in the header were well-formed, and parsed the
1447 element content after inspecting only the first char of the tag.
1448 Assumptions at that stage could crash the receive process on malformed
1451 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1452 While running the DKIM ACL we operate on the Permanent memory pool so that
1453 variables created with "set" persist to the DATA ACL. Also (at any time)
1454 DNS lookups that fail create cache records using the Permanent pool. But
1455 expansions release any allocations made on the current pool - so a dnsdb
1456 lookup expansion done in the DKIM ACL releases the memory used for the
1457 DNS negative-cache, and bad things result. Solution is to switch to the
1458 Main pool for expansions.
1459 While we're in that code, add checks on the DNS cache during store_reset,
1460 active in the testsuite.
1461 Problem spotted, and debugging aided, by Wolfgang Breyha.
1463 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1464 When none of the hosts presented to a transport match an already-open
1465 connection, close it and proceed with the list. Previously we would
1466 queue the message. Spotted by Lena with Yahoo, probably involving
1469 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1470 Previously a spurious "250 OK id=" response was appended to the proper
1473 JH/08 The "support for" informational output now, which built with Content
1474 Scanning support, has a line for the malware scanner interfaces compiled
1475 in. Interface can be individually included or not at build time.
1477 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1478 by the template makefile "src/EDITME". The "STREAM" support for an older
1479 ClamAV interface method is removed.
1481 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1482 rows affected is given instead).
1484 JH/11 The runtime Berkeley DB library version is now additionally output by
1485 "exim -d -bV". Previously only the compile-time version was shown.
1487 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1488 SMTP connection. Previously, when one had more recipients than the
1489 first, an abortive onward connection was made. Move to full support for
1490 multiple onward connections in sequence, handling cutthrough connection
1491 for all multi-message initiating connections.
1493 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1494 routers. Previously, a multi-recipient message would fail to match the
1495 onward-connection opened for the first recipient, and cause its closure.
1497 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1498 a timeout on read on a GnuTLS initiating connection, resulting in the
1499 initiating connection being dropped. This mattered most when the callout
1500 was marked defer_ok. Fix to keep the two timeout-detection methods
1503 JH/15 Relax results from ACL control request to enable cutthrough, in
1504 unsupported situations, from error to silently (except under debug)
1505 ignoring. This covers use with PRDR, frozen messages, queue-only and
1508 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1510 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1511 metadata, resulting in a crash in free().
1513 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1514 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1515 Broken also in d185889f4, with init system revamp.
1517 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1518 Previously we abruptly closed the connection after reading a malware-
1519 found indication; now we go on to read the "scan ok" response line,
1522 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1523 ACL. Previously, a crash would result.
1525 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1526 macro text after a replacement (previously it was only once per line) and
1527 by skipping builtin macros when searching for an uppercase lead character.
1529 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1530 for the build is renamed.
1532 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1533 was allocated for every new TLS startup, meaning one per message. Fix
1534 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1536 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1537 reported the original. Fix to report (as far as possible) the ACL
1538 result replacing the original.
1540 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1541 OpenSSL. Certificate information is loaded for every new TLS startup,
1542 and the resources needed to be freed.
1544 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1546 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1549 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1550 DATA response info to the (existing) per-recipient response info for
1551 the "C=" log element. It can have useful tracking info from the
1552 destination system. Patch from Simon Arlott.
1554 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1555 length value. Previously this would segfault.
1557 HS/02 Support Avast multiline protoocol, this allows passing flags to
1558 newer versions of the scanner.
1560 JH/28 Ensure that variables possibly set during message acceptance are marked
1561 dead before release of memory in the daemon loop. This stops complaints
1562 about them when the debug_store option is enabled. Discovered specifically
1563 for sender_rate_period, but applies to a whole set of variables.
1564 Do the same for the queue-runner and queue-list loops, for variables set
1565 from spool message files. Do the same for the SMTP per-message loop, for
1566 certain variables indirectly set in ACL operations.
1568 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1569 as a multi-recipient message from a mailinglist manager). The coding had
1570 an arbitrary cutoff number of characters while checking for more input;
1571 enforced by writing a NUL into the buffer. This corrupted long / fast
1572 input. The problem was exposed more widely when more pipelineing of SMTP
1573 responses was introduced, and one Exim system was feeding another.
1574 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1575 receiving system, and refused recipients seen by the sending system
1576 (propating to people being dropped from mailing lists).
1577 Discovered and pinpointed by David Carter.
1579 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1580 replaced by the ${authresults } expansion.
1582 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1584 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1585 allows proper process termination in container environments.
1587 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1588 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1590 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1591 and "err_perm", deprecated since 4.83 when the RFC-defined words
1592 "temperror" and "permerror" were introduced.
1594 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1595 transport-filters or DKIM-signing. The restriction was lost in the
1596 consolidation of verify-callout and delivery SMTP handling.
1597 Extend the restriction to also cover ARC-signing.
1599 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1600 in defer=pass mode supply a 450 to the initiator. Previously the message
1603 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1604 tls_require_ciphers is used as before.
1606 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1607 "pass_unscanned". Only tmpfails from the scanner are written to
1608 the paniclog, as they may require admin intervention (permission
1609 denied, license issues). Other scanner errors (like decompression
1610 bombs) do not cause a paniclog entry.
1612 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1613 Previously it was possible to log spurious information in receive log
1616 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1617 triggered odd behaviour from Outlook Express clients.
1619 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1622 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1623 since the IETF WG has not yet settled on that versus the original
1624 "bare" representation.
1626 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1627 Previously the millisecond value corrupted the output.
1628 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1629 corrupted the output.
1635 JH/01 Rework error string handling in TLS interface so that the caller in
1636 more cases is responsible for logging. This permits library-sourced
1637 string to be attached to addresses during delivery, and collapses
1638 pairs of long lines into single ones.
1640 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1641 during configuration. Wildcards are allowed and expanded.
1643 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1644 This permits better logging.
1646 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1647 do not close it down and have the child transport start it up again on
1648 the passed-on TCP connection. Instead, proxy the child (and any
1649 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1650 affected: the continued delivery log lines do not have any DNSSEC, TLS
1651 Certificate or OCSP information. TLS cipher information is still logged.
1653 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1654 identical IP addresses on different listening ports. Will also affect
1657 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1658 add noisy ifdef guards to special-case this sillyness.
1659 Patch from Bernd Kuhls.
1661 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1662 than 255 are no longer allowed.
1664 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1665 Disable the session-cache too, which might reduce our load. Since we
1666 currrectly use a new context for every connection, both as server and
1667 client, there is no benefit for these.
1668 GnuTLS appears to not support tickets server-side by default (we don't
1669 call gnutls_session_ticket_enable_server()) but client side is enabled
1670 by default on recent versions (3.1.3 +) unless the PFS priority string
1673 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1674 <https://reproducible-builds.org/specs/source-date-epoch/>.
1676 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1677 the check for any unsuccessful recipients did not notice the limit, and
1678 erroneously found still-pending ones.
1680 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1681 MSG_MORE. Only in-clear (not on TLS connections).
1683 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1684 transport-filter is involved we can buffer the headers in memory for
1685 creating the signature, and read the spool data file once for the
1686 signature and again for transmission.
1688 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1689 4.77 as the kernel support then wasn't solid, having issues in 64bit
1690 mode. Now, it's been long enough. Add support for FreeBSD also.
1692 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1693 case where the routing stage had gathered several addresses to send to
1694 a host before calling the transport for the first, we previously failed
1695 to close down TLS in the old transport process before passing the TCP
1696 connection to the new process. The new one sent a STARTTLS command
1697 which naturally failed, giving a failed delivery and bloating the retry
1698 database. Investigation and fix prototype from Wolfgang Breyha.
1700 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1701 false-negatives in the check that the sender had not preempted a response
1702 or prompt from Exim (running as a server), due to that code's lack of
1703 awareness of the SMTP input buffering.
1705 PP/04 Add commandline_checks_require_admin option.
1706 Exim drops privileges sanely, various checks such as -be aren't a
1707 security problem, as long as you trust local users with access to their
1708 own account. When invoked by services which pass untrusted data to
1709 Exim, this might be an issue. Set this option in main configuration
1710 AND make fixes to the calling application, such as using `--` to stop
1713 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1714 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1715 responses to those, into a single TLS record each way (this usually means
1716 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1719 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1720 affects you only if you're dancing at the edge of the param size limits.
1721 If you are, and this message makes sense to you, then: raise the
1722 configured limit or use OpenSSL 1.1. Nothing we can do for older
1725 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1726 cmdline element to get the documented default one. Previously it was
1729 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1730 are unsuable for non-message-specific reasons - eg. connection timeout,
1731 banner-time rejection.
1733 JH/16 Fix logging of delivery remote port, when specified by router, under
1736 PP/06 Repair manualroute's ability to take options in any order, even if one
1737 is the name of a transport.
1740 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1742 JH/17 Change the list-building routines interface to use the expanding-string
1743 triplet model, for better allocation and copying behaviour.
1745 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1746 Previously it was constructed the first time a possibly-matching string
1747 was met in the configuration file input during startup; now it is done
1750 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1751 compatible one, to avoid the (poorly documented) possibility of a config
1752 file in the working directory redirecting the DB files, possibly correpting
1753 some existing file. CVE-2017-10140 assigned for BDB.
1755 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1756 cache-hot. Previously, although the result was properly cached, the
1757 initial verify call returned a defer.
1759 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1760 the main verify for receipient in uncached-mode.
1762 JH/22 Retire historical build files to an "unsupported" subdir. These are
1763 defined as "ones for which we have no current evidence of testing".
1765 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1766 if present. Previously it was ignored.
1768 JH/24 Start using specified-initialisers in C structure init coding. This is
1769 a C99 feature (it's 2017, so now considered safe).
1771 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1772 if was a fixed-sized field and bitmask ops via macros; it is now more
1775 PP/07 GitHub PR 56: Apply MariaDB build fix.
1776 Patch provided by Jaroslav Škarvada.
1778 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1779 during Coverity cleanups [4.87 JH/47]
1780 Diagnosis and fix provided by Michael Fischer v. Mollard.
1782 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1783 the right size to place the terminating semicolon on its own folded
1784 line, the header hash was calculated to an incorrect value thanks to
1785 the (relaxed) space the fold became.
1787 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1788 and confused the parent.
1790 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1791 which could crash as a result. This could lead to undeliverable messages.
1793 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1796 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1797 matters on fast-turnover and PID-randomising systems, which were getting
1798 out-of-order delivery.
1800 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1801 a possibly-overlapping copy. The symptom was that "Remote host closed
1802 connection in response to HELO" was logged instead of the actual 4xx
1805 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1806 Previously only that bufferd was discarded, resulting in SYMTP command
1809 JH/32 DKIM: when a message has multiple signatures matching an identity given
1810 in dkim_verify_signers, run the dkim acl once for each. Previously only
1811 one run was done. Bug 2189.
1813 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1814 "panic the current process" to "deliberately defer". The panic log is
1815 still written with the problem list name; the mail and reject logs now
1816 get a temp-reject line for the message that was being handled, saying
1817 something like "domains check lookup or other defer". The SMTP 451
1818 message is still "Temporary local problem".
1820 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1821 A crafted sequence of BDAT commands could result in in-use memory beeing
1822 freed. CVE-2017-16943.
1824 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1825 from SMTP input. Previously it was always done; now only done for DATA
1826 and not BDAT commands. CVE-2017-16944.
1828 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1829 to the message (such as an overlong header line). Previously this was
1830 not done and we did not exit BDAT mode. Followon from the previous item
1831 though a different problem.
1837 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1838 than -2003 did; needs libidn2 in addition to libidn.
1840 JH/02 The path option on a pipe transport is now expanded before use.
1842 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1843 Patch provided by "Björn", documentation fix added too.
1845 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1846 missing a wire-to-host endian conversion.
1848 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1849 close after a BDAT command line could be taken as a following command,
1850 giving a synch failure. Fix by only checking for synch immediately
1851 before acknowledging the chunk.
1853 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1854 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1855 macro. Patches provided by Josh Soref.
1857 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1858 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1859 by IANA as a service extension. However, John Klensin suggests that we
1862 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1863 the dkim code may be unix-mode line endings rather than smtp wire-format
1864 CRLF, so prepend a CR to any bare LF.
1866 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1867 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1869 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1870 the first were themselves being wrongly included in the feed into dkim
1871 processing; with most chunk sizes in use this resulted in an incorrect
1872 body hash calculated value.
1874 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1875 DKIM signature block, for verification. Although advised against by
1876 standards it is specifically not ruled illegal.
1878 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1880 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1881 missing a body hash (the bh= tag).
1883 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1884 It seems that HAProxy sends the Proxy Protocol information in clear and
1885 only then does a TLS startup, so do the same.
1887 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1888 TCP connections (such as for Spamd) unless the daemon successfully set
1889 Fast Open mode on its listening sockets. This fixes breakage seen on
1890 too-old kernels or those not configured for Fast Open, at the cost of
1891 requiring both directions being enabled for TFO, and TFO never being used
1892 by non-daemon-related Exim processes.
1894 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1895 endings, at least on the first header line. Try to canonify any that get
1896 past that check, despite the cost.
1898 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1899 now limited to an arbitrary five deep, while parsing addresses with the
1900 strip_excess_angle_brackets option enabled.
1902 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1903 instead leave the unprompted TLS handshake in socket buffer for the
1904 TLS library to consume.
1906 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1908 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1910 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1911 region in one operation, for speed, and this leaves assigned pointers
1912 dangling. Add checks run only under the testsuite which checks all
1913 variables at a store-reset and panics on a dangling pointer; add code
1914 explicitly nulling out all the variables discovered. Fixes one known
1915 bug: a transport crash, where a dangling pointer for $sending_ip_address
1916 originally assigned in a verify callout, is re-used.
1918 PP/06 Drop '.' from @INC in various Perl scripts.
1920 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1922 PP/08 Reduce a number of compilation warnings under clang; building with
1923 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1924 should be warning-free.
1926 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1928 HS/01 Fix portability problems introduced by PP/08 for platforms where
1929 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1931 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1932 chunk. This allows us to accept broken chunked messages. We need a more
1933 general solution here.
1935 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1936 already-broken messages in the queue.
1938 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1940 JH/19 Fix reference counting bug in routing-generated-address tracking.
1946 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1947 supports it and a size is available (ie. the sending peer gave us one).
1949 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1950 years of being deprecated). The replacements are the ACLs
1951 acl_smtp_mime and acl_not_smtp_mime.
1953 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1954 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1955 or even an in-clear connection were permitted. Now, if the host lookup
1956 was dnssec and dane was requested then the host is only used if the
1957 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1958 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1959 if one fails this test.
1960 This means that a poorly-configured remote DNS will make it incommunicado;
1961 but it protects against a DNS-interception attack on it.
1963 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1964 non-noisy when a race steals the message being considered.
1966 JH/05 If main configuration option tls_certificate is unset, generate a
1967 self-signed certificate for inbound TLS connections.
1969 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1970 in rewrites and routers.
1972 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1973 and logged a warning sing 4.83; now they are a configuration file error.
1975 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1976 (lacking @domain). Apply the same qualification processing as RCPT.
1978 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1980 JH/10 Support ${sha256:} applied to a string (as well as the previous
1983 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1984 a cutthrough deliver is pending, as we always want to make a connection.
1985 This also avoids re-routing the message when later placing the cutthrough
1986 connection after a verify cache hit.
1987 Do not update it with the verify result either.
1989 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1990 when routing results in more than one destination address.
1992 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1993 signing (which inhibits the cutthrough capability). Previously only
1994 the presence of an option was tested; now an expansion evaluating as
1995 empty is permissible (obviously it should depend only on data available
1996 when the cutthrough connection is made).
1998 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1999 the relevant preceding SMTP command did not note the pipelining mode.
2001 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2002 Previously they were not counted.
2004 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2005 as one having no matching records. Previously we deferred the message
2006 that needed the lookup.
2008 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2009 distinguished as "(=".
2011 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2012 for missing MX records. Previously it only worked for missing A records.
2014 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2016 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2017 after the data-go-ahead and data-ack. Patch from Jason Betts.
2019 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2020 even for a "none" policy. Patch from Tony Meyer.
2022 JH/22 Fix continued use of a connection for further deliveries. If a port was
2023 specified by a router, it must also match for the delivery to be
2026 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2027 When one of the recipients of a message was unsuitable for the connection
2028 (has no matching addresses), we lost track of needing to mark it
2029 deferred. As a result mail would be lost.
2031 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2033 JH/25 Decoding ACL controls is now done using a binary search; the source code
2034 takes up less space and should be simpler to maintain. Merge the ACL
2035 condition decode tables also, with similar effect.
2037 JH/26 Fix problem with one_time used on a redirect router which returned the
2038 parent address unchanged. A retry would see the parent address marked as
2039 delivered, so not attempt the (identical) child. As a result mail would
2042 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2043 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2044 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2047 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2048 Main config options check_{log,spool}_{inodes,space} are now
2049 100 inodes, 10MB unless set otherwise in the configuration.
2051 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2052 Previously it only applied to the main-section connection policy
2055 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2057 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2058 by me. Added RFC7919 DH primes as an alternative.
2060 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2061 are not in the system include path.
2063 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2064 GnuTLS, when a session startup failed (eg because the client disconnected)
2065 Exim did stdio operations after fclose. This was exposed by a recent
2066 change which nulled out the file handle after the fclose.
2068 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2069 signed directly by the cert-signing cert, rather than an intermediate
2070 OCSP-signing cert. This is the model used by LetsEncrypt.
2072 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2074 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2075 an incoming connection.
2077 HS/02 Bug 1802: Do not half-close the connection after sending a request
2080 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2081 fallback to "prime256v1".
2083 JH/34 SECURITY: Use proper copy of DATA command in error message.
2084 Could leak key material. Remotely exploitable. CVE-2016-9963.
2090 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2091 and 3.4.4 - once the server is enabled to respond to an OCSP request
2092 it does even when not requested, resulting in a stapling non-aware
2093 client dropping the TLS connection.
2095 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2096 support variable-length bit vectors. No functional change.
2098 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2099 The I= interface field on outgoing lines is now after the H= remote
2100 host field, same as incoming lines. There is a separate
2101 outgoing_interface log selector which allows you to disable the
2104 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2105 If not running log_selector +smtp_connection the mainlog would be held
2106 open indefinitely after a "too many connections" event, including to a
2107 deleted file after a log rotate. Leave the per net connection logging
2108 leaving it open for efficiency as that will be quickly detected by the
2109 check on the next write.
2111 HS/01 Bug 1671: Fix post transport crash.
2112 Processing the wait-<transport> messages could crash the delivery
2113 process if the message IDs didn't exist for some reason. When
2114 using 'split_spool_directory=yes' the construction of the spool
2115 file name failed already, exposing the same netto behaviour.
2117 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2118 mime_regex ACL conditions.
2120 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2121 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2122 to HELO, local diagnostic string.
2124 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2125 log line to debug. Even when configured with a tls authenticator many
2126 client connections are expected to not authenticate in this way, so
2127 an authenticate fail is not an error.
2129 HS/02 Add the Exim version string to the process info. This way exiwhat
2130 gives some more detail about the running daemon.
2132 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2133 matter for fast-change records such as DNSBLs.
2135 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2136 retry record, even if constant. There may be multiple transports with
2137 different interface settings and the retry behaviour needs to be kept
2140 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2142 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2144 JH/10 Bug 840: fix log_defer_output option of pipe transport
2146 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2147 hosts_randomize. This matters a lot when combined with mua_wrapper.
2149 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2150 ${quote_pgsql:<string>} operator.
2152 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2154 JH/14 Change header file name for freeradius-client. Relevant if compiling
2155 with Radius support; from the Gentoo tree and checked under Fedora.
2157 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2159 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2160 option transport_filter, meaning no filtering.
2162 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2164 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2165 defaults to "*" (all hosts). The variable is now available when not built
2166 with TLS, default unset, mainly to enable keeping the testsuite sane.
2167 If a server certificate is not supplied (via tls_certificate) an error is
2168 logged, and clients will find TLS connections fail on startup. Presumably
2169 they will retry in-clear.
2170 Packagers of Exim are strongly encouraged to create a server certificate
2171 at installation time.
2173 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2174 with the $config_file variable.
2176 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2177 in transport context, after the attempt, and per-recipient. The latter type
2178 is per host attempted. The event data is the error message, and the errno
2179 information encodes the lookup type (A vs. MX) used for the (first) host,
2180 and the trailing two digits of the smtp 4xx response.
2182 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2183 to write to mainlog (or rejectlog, paniclog) in the window between file
2184 creation and permissions/ownership being changed. Particularly affects
2185 installations where exicyclog is run as root, rather than exim user;
2186 result is that the running daemon panics and dies.
2188 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2190 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2191 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2192 "pri" and "weight". Note that the previous implicit priority given by the
2193 list order is no longer honoured.
2195 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2196 for DKIM processing.
2198 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2199 by defining SUPPORT_SOCKS.
2201 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2202 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2203 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2204 variables are renamed to proxy_{local,external}_{address,port}.
2206 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2207 for a build by defining SUPPORT_I18N
2209 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2210 of the query string, and make ${quote_redis:} do that quoting.
2212 JH/29 Move Events support from Experimental to mainline, enabled by default
2213 and removable for a build by defining DISABLE_EVENT.
2215 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2217 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2218 cached by the daemon.
2220 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2221 by defining LOOKUP_REDIS. The libhiredis library is required.
2223 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2224 keys are given for lookup.
2226 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2227 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2228 only supported when built with TLS support. The PolarSSL SHA routines
2229 are still used when the TLS library is too old for convenient support.
2231 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2232 openssl_options), for security. OpenSSL forces this from version 1.1.0
2233 server-side so match that on older versions.
2235 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2236 allocation for $value could be released as the expansion processing
2237 concluded, but leaving the global pointer active for it.
2239 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2240 and to use the domains and local_parts ACL conditions.
2242 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2243 incorrectly not doubled on cutthrough transmission, hence seen as a
2244 body-termination at the receiving system - resulting in truncated mails.
2245 Commonly the sender saw a TCP-level error, and retransmitted the message
2246 via the normal store-and-forward channel. This could result in duplicates
2247 received - but deduplicating mailstores were liable to retain only the
2248 initial truncated version.
2250 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2252 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2254 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2255 we're in there, support oversigning also; bug 1309.
2257 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2259 HS/04 Add support for keep_environment and add_environment options.
2261 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2262 either intentional arithmetic overflow during PRNG, or testing config-
2265 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2266 delivery resulted in actual delivery. Cancel cutthrough before DATA
2269 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2270 rejects a recipient: pass the reject to the originator.
2272 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2273 Many were false-positives and ignorable, but it's worth fixing the
2276 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2277 for the new environment-manipulation done at startup. Move the routines
2278 from being local to tls.c to being global via the os.c file.
2280 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2281 an extract embedded as result-arg for a map, the first arg for extract
2282 is unavailable so we cannot tell if this is a numbered or keyed
2283 extraction. Accept either.
2289 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2292 JH/02 The smtp transport option "multi_domain" is now expanded.
2294 JH/03 The smtp transport now requests PRDR by default, if the server offers
2297 JH/04 Certificate name checking on server certificates, when exim is a client,
2298 is now done by default. The transport option tls_verify_cert_hostnames
2299 can be used to disable this per-host. The build option
2300 EXPERIMENTAL_CERTNAMES is withdrawn.
2302 JH/05 The value of the tls_verify_certificates smtp transport and main options
2303 default to the word "system" to access the system default CA bundle.
2304 For GnuTLS, only version 3.0.20 or later.
2306 JH/06 Verification of the server certificate for a TLS connection is now tried
2307 (but not required) by default. The verification status is now logged by
2308 default, for both outbound TLS and client-certificate supplying inbound
2311 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2314 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2315 Status Notification (bounce) messages are now MIME format per RFC 3464.
2316 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2317 under the control of the dsn_advertise_hosts option, and routers may
2318 have a dsn_lasthop option.
2320 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2321 default, modifiable by a malware= option. The list separator for
2322 the options can now be changed in the usual way. Bug 68.
2324 JH/10 The smtp_receive_timeout main option is now expanded before use.
2326 JH/11 The incoming_interface log option now also enables logging of the
2327 local interface on delivery outgoing connections.
2329 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2330 if the interface and destination host and port all match.
2332 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2335 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2336 Patch from Andrew Lewis.
2338 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2339 now supports optional time-restrictions, weighting, and priority
2340 modifiers per server. Patch originally by <rommer@active.by>.
2342 JH/16 The spamd_address main option now supports a mixed list of local
2343 and remote servers. Remote servers can be IPv6 addresses, and
2344 specify a port-range.
2346 JH/17 Bug 68: The spamd_address main option now supports an optional
2347 timeout value per server.
2349 JH/18 Bug 1581: Router and transport options headers_add/remove can
2350 now have the list separator specified.
2352 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2355 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2358 JH/21 Support for the A6 type of dns record is withdrawn.
2360 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2361 rather than the verbs used.
2363 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2364 from 255 to 1024 chars.
2366 JH/24 Verification callouts now attempt to use TLS by default.
2368 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2369 are generic router options now. The defaults didn't change.
2371 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2372 Original patch from Alexander Shikoff, worked over by JH.
2374 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2375 files if ZCAT_COMMAND is not executable.
2377 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2379 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2381 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2382 Normally benign, it bites when the pair was led to by a CNAME;
2383 modern usage is to not canonicalize the domain to a CNAME target
2384 (and we were inconsistent anyway for A-only vs AAAA+A).
2386 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2388 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2389 when evaluating $sender_host_dnssec.
2391 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2392 $sender_helo_dnssec variable.
2394 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2396 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2398 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2400 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2401 documented as working, but never had. Support all but $spam_report.
2403 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2404 added for tls authenticator.
2406 HS/03 Add perl_taintmode main config option
2412 TL/01 When running the test suite, the README says that variables such as
2413 no_msglog_check are global and can be placed anywhere in a specific
2414 test's script, however it was observed that placement needed to be near
2415 the beginning for it to behave that way. Changed the runtest perl
2416 script to read through the entire script once to detect and set these
2417 variables, reset to the beginning of the script, and then run through
2418 the script parsing/test process like normal.
2420 TL/02 The BSD's have an arc4random API. One of the functions to induce
2421 adding randomness was arc4random_stir(), but it has been removed in
2422 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2423 function when detected.
2425 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2426 cause callback expansion.
2428 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2429 syntax errors in an expansion can be treated as a string instead of
2430 logging or causing an error, due to the internal use of bool_lax
2431 instead of bool when processing it.
2433 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2434 server certificates when making smtp deliveries.
2436 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2438 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2440 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2442 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2443 Merged patch from Sebastian Wiedenroth.
2445 JH/05 Fix results-pipe from transport process. Several recipients, combined
2446 with certificate use, exposed issues where response data items split
2447 over buffer boundaries were not parsed properly. This eventually
2448 resulted in duplicates being sent. This issue only became common enough
2449 to notice due to the introduction of connection certificate information,
2450 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2452 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2453 size buffer was used, resulting in syntax errors when an expansion
2456 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2457 version 3.3.6 or later.
2459 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2460 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2461 both become "event_action", the variables become $event_name, $event_data
2462 and $event_defer_errno. There is a new variable $verify_mode, usable in
2463 routers, transports and related events. The tls:cert event is now also
2464 raised for inbound connections, if the main configuration event_action
2467 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2468 early OCSP support, but no stapling (appears to be less than 1.0.0).
2470 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2471 server certificate names available under the smtp transport option
2472 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2475 JH/10 Time-related extraction expansions from certificates now use the main
2476 option "timezone" setting for output formatting, and are consistent
2477 between OpenSSL and GnuTLS compilations. Bug 1541.
2479 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2480 encoded parameter in the incoming message. Bug 1558.
2482 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2483 include certificate info, eximon was claiming there were spoolfile
2486 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2488 JH/14 Log delivery-related information more consistently, using the sequence
2489 "H=<name> [<ip>]" wherever possible.
2491 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2492 are problematic for Debian distribution, omit them from the release
2495 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2497 JH/16 Fix string representation of time values on 64bit time_t architectures.
2500 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2501 output list separator was used.
2506 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2507 checkers that were complaining about end of non-void function with no
2510 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2511 This was a regression introduced in 4.83 by another bugfix.
2513 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2515 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2516 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2522 TF/01 Correctly close the server side of TLS when forking for delivery.
2524 When a message was received over SMTP with TLS, Exim failed to clear up
2525 the incoming connection properly after forking off the child process to
2526 deliver the message. In some situations the subsequent outgoing
2527 delivery connection happened to have the same fd number as the incoming
2528 connection previously had. Exim would try to use TLS and fail, logging
2529 a "Bad file descriptor" error.
2531 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2532 utilities have not been installed.
2534 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2535 temporary space as the ACL may create new global variables.
2537 TL/01 LDAP support uses per connection or global context settings, depending
2538 upon the detected version of the libraries at build time.
2540 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2541 to extract and use the src ip:port in logging and expansions as if it
2542 were a direct connection from the outside internet. PPv2 support was
2543 updated based on HAProxy spec change in May 2014.
2545 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2547 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2548 Properly escape header and check for NULL return.
2550 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2553 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2555 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2556 characters in header names, implemented as a verify condition.
2557 Contributed by Michael Fischer v. Mollard.
2559 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2560 results permerror and temperror. Previous values are deprecated but
2561 still accepted. In a future release, err_perm and err_temp will be
2562 completely removed, which will be a backward incompatibility if the
2563 ACL tests for either of these two old results. Patch contributed by
2564 user bes-internal on the mailing list.
2566 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2568 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2569 selectors, in both main and reject logs.
2571 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2574 JH/07 Add malware type "sock" for talking to simple daemon.
2576 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2578 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2579 routers/transports under cutthrough routing.
2581 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2582 numbers. Touch up "bool" conditional to keep the same definition.
2584 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2586 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2588 JH/12 Expand items in router/transport headers_add or headers_remove lists
2589 individually rather than the list as a whole. Bug 1452.
2591 Required for reasonable handling of multiple headers_ options when
2592 they may be empty; requires that headers_remove items with embedded
2593 colons must have them doubled (or the list-separator changed).
2595 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2596 view the policy declared in the DMARC record. Currently, $dmarc_status
2597 is a combined value of both the record presence and the result of the
2600 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2602 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2603 dnslookup router and the smtp transport (applying to the forward
2606 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2607 of ldap servers used for a specific lookup. Patch provided by Heiko
2610 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2611 New variable $lookup_dnssec_authenticated for observability.
2613 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2614 Patch submitted by Lars Timman.
2616 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2618 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2619 Requires trusted mode and valid format message id, aborts otherwise.
2620 Patch contributed by Heiko Schlichting.
2622 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2623 certextract with support for various fields. Bug 1358.
2625 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2626 is requested by default, modifiable by smtp transport option
2629 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2630 operate on certificate variables to give certificate fingerprints
2631 Also new ${sha256:cert_variable}.
2633 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2635 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2636 Christian Aistleitner.
2638 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2640 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2641 file. Patch from Wolfgang Breyha.
2643 JH/25 Expand the coverage of the delivery $host and $host_address to
2644 client authenticators run in verify callout. Bug 1476.
2646 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2647 align with daemon_smtp_ports. Bug 72.
2649 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2650 support and error reporting did not work properly.
2652 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2653 and is readable. Patch from Andrew Colin Kissa.
2655 TL/14 Enhance documentation of ${run expansion and how it parses the
2656 commandline after expansion, particularly in the case when an
2657 unquoted variable expansion results in an empty value.
2659 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2661 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2662 by renaming away from it. Addresses GH issue 15, reported by
2665 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2667 TL/15 SECURITY: prevent double expansion in math comparison functions
2668 (can expand unsanitized data). Not remotely exploitable.
2675 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2677 PP/02 Make -n do something, by making it not do something.
2678 When combined with -bP, the name of an option is not output.
2680 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2683 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2684 $sender_host_name and config options to manage this, and basic check
2687 PP/05 DSCP support for outbound connections and control modifier for inbound.
2689 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2690 (Only plugin which currently uses this is kerberos4, which nobody should
2691 be using, but we should make it available and other future plugins might
2692 conceivably use it, even though it would break NAT; stuff *should* be
2693 using channel bindings instead).
2695 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2696 name; added for Sendmail compatibility; requires admin caller.
2697 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2698 just ignore it); requires trusted caller.
2699 Also parse but ignore: -Ac -Am -X<logfile>
2702 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2704 TL/02 Add +smtp_confirmation as a default logging option.
2706 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2707 Patch by Magnus Holmgren from 2007-02-20.
2709 TL/04 Bugzilla 1281 - Spec typo.
2710 Bugzilla 1283 - Spec typo.
2711 Bugzilla 1290 - Spec grammar fixes.
2713 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2715 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2717 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2718 mailing list by by Dmitry Isaikin.
2720 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2722 JH/02 Support "G" suffix to numbers in ${if comparisons.
2724 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2726 NM/01 Bugzilla 1197 - Spec typo
2727 Bugzilla 1196 - Spec examples corrections
2729 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2731 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2732 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2735 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2736 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2738 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2739 "acl {{name}{arg}...}", and optional args on acl condition
2742 JH/05 Permit multiple router/transport headers_add/remove lines.
2744 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2746 JH/07 Avoid using a waiting database for a single-message-only transport.
2747 Performance patch from Paul Fisher. Bugzilla 1262.
2749 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2752 JH/09 Add $headers_added variable, with content from use of ACL modifier
2753 add_header (but not yet added to the message). Bugzilla 199.
2755 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2756 Pulled from Bugzilla 817 by Wolfgang Breyha.
2758 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2760 (nb: this is the same fix as in Exim 4.80.1)
2762 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2765 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2768 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2770 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2771 advertise SMTP AUTH mechanism to us, instead of a generic
2772 protocol violation error. Also, make Exim more robust to bad
2773 data from the Dovecot auth socket.
2775 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2777 When a queue runner is handling a message, Exim first routes the
2778 recipient addresses, during which it prunes them based on the retry
2779 hints database. After that it attempts to deliver the message to
2780 any remaining recipients. It then updates the hints database using
2783 So if a recipient address works intermittently, it can get repeatedly
2784 deferred at routing time. The retry hints record remains fresh so the
2785 address never reaches the final cutoff time.
2787 This is a fairly common occurrence when a user is bumping up against
2788 their storage quota. Exim had some logic in its local delivery code
2789 to deal with this. However it did not apply to per-recipient defers
2790 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2792 This change adds a proper retry rule check during routing so that the
2793 final cutoff time is checked against the message's age. We only do
2794 this check if there is an address retry record and there is not a
2795 domain retry record; this implies that previous attempts to handle
2796 the address had the retry_use_local_parts option turned on. We use
2797 this as an approximation for the destination being like a local
2798 delivery, as in LMTP.
2800 I suspect this new check makes the old local delivery cutoff check
2801 redundant, but I have not verified this so I left the code in place.
2803 TF/02 Correct gecos expansion when From: is a prefix of the username.
2805 Test 0254 submits a message to Exim with the header
2809 When I ran the test suite under the user fanf2, Exim expanded
2810 the header to contain my full name, whereas it should have added
2811 a Resent-Sender: header. It erroneously treats any prefix of the
2812 username as equal to the username.
2814 This change corrects that bug.
2816 GF/01 DCC debug and logging tidyup
2817 Error conditions log to paniclog rather than rejectlog.
2818 Debug lines prefixed by "DCC: " to remove any ambiguity.
2820 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2822 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2823 Bug spotted by Jeremy Harris; was flawed since initial commit.
2824 Would have resulted in OCSP responses post-SNI triggering an Exim
2825 NULL dereference and crash.
2827 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2829 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2830 Bug detection, analysis and fix by Samuel Thibault.
2831 Bugzilla 1331, Debian bug #698092.
2833 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2835 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2836 Server implementation by Todd Lyons, client by JH.
2837 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2838 config variable "prdr_enable" controls whether the server
2839 advertises the facility. If the client requests PRDR a new
2840 acl_data_smtp_prdr ACL is called once for each recipient, after
2841 the body content is received and before the acl_smtp_data ACL.
2842 The client is controlled by both of: a hosts_try_prdr option
2843 on the smtp transport, and the server advertisement.
2844 Default client logging of deliveries and rejections involving
2845 PRDR are flagged with the string "PRDR".
2847 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2848 fclose(). Diagnosis by Todd Lyons.
2850 PP/17 Update configure.default to handle IPv6 localhost better.
2851 Patch by Alain Williams (plus minor tweaks).
2854 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2855 This is now consistent with GnuTLS, and is now documented: the
2856 previous undocumented portable approach to treating the option as
2857 unset was to force an expansion failure. That still works, and
2858 an empty string is now equivalent.
2860 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2861 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2862 not performing validation itself.
2864 PP/20 Added force_command boolean option to pipe transport.
2865 Patch from Nick Koston, of cPanel Inc.
2867 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2870 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2872 PP/21 Fix eximon continuous updating with timestamped log-files.
2873 Broken in a format-string cleanup in 4.80, missed when I repaired the
2874 other false fix of the same issue.
2875 Report and fix from Heiko Schlichting.
2878 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2879 Report from Prashanth Katuri.
2881 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2882 It's SecureTransport, so affects any MacOS clients which use the
2883 system-integrated TLS libraries, including email clients.
2885 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2886 using a MIME ACL for non-SMTP local injection.
2887 Report and assistance in diagnosis by Warren Baker.
2889 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2891 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2893 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2894 last id that failed so it may be referenced in subsequent ACL's.
2896 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2899 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2900 ldap library initialization, allowing self-signed CA's to be
2901 used. Also properly sets require_cert option later in code by
2902 using NULL (global ldap config) instead of ldap handle (per
2903 session). Bug diagnosis and testing by alxgomz.
2905 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2906 the src/util/ subdirectory.
2908 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2909 renamed to Transport Post Delivery Action by Jeremy Harris, as
2912 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2913 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2914 redis_servers = needs to be configured which will be used by the redis
2915 lookup. Patch from Warren Baker, of The Packet Hub.
2917 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2919 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2920 hostname or reverse DNS when processing a host list. Used suggestions
2921 from multiple comments on this bug.
2923 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2925 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2926 Missed a few lines, added it to make the runtest require no keyboard
2929 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2930 contains upper case chars. Make router use caseful_local_part.
2932 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2933 support when GnuTLS has been built with p11-kit.
2939 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2941 This, or similar/improved, will also be change PP/11 of 4.82.
2947 PP/01 Handle short writes when writing local log-files.
2948 In practice, only affects FreeBSD (8 onwards).
2949 Bugzilla 1053, with thanks to Dmitry Isaikin.
2951 NM/01 Bugzilla 949 - Documentation tweak
2953 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2956 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2958 PP/02 Implemented gsasl authenticator.
2960 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2962 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2963 `pkg-config foo` for cflags/libs.
2965 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2966 with rest of GSASL and with heimdal_gssapi.
2968 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2969 `pkg-config foo` for cflags/libs for the TLS implementation.
2971 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2972 properties get this fed in as external SSF. A number of robustness
2973 and debugging improvements to the cyrus_sasl authenticator.
2975 PP/08 cyrus_sasl server now expands the server_realm option.
2977 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2978 Patch by Jeremy Harris.
2980 PP/10 Added dbmjz lookup type.
2982 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2984 PP/12 MAIL args handles TAB as well as SP, for better interop with
2985 non-compliant senders.
2986 Analysis and variant patch by Todd Lyons.
2988 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2989 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2990 Patch from Dirk Mueller <dmueller@suse.com>
2992 PP/13 tls_peerdn now print-escaped for spool files.
2993 Observed some $tls_peerdn in wild which contained \n, which resulted
2994 in spool file corruption.
2996 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2997 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2998 or write after TLS renegotiation, which otherwise led to messages
3001 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3002 as a tracking header (ie: a signed header comes before the signature).
3003 Patch from Wolfgang Breyha.
3005 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3006 comma-sep list; embedded commas doubled.
3008 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3010 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3012 Report and patch from Dmitry Banschikov.
3014 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3015 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3016 failures. We appear to now support TLS1.1+ with Exim.
3018 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3019 lets Exim select keys and certificates based upon TLS SNI from client.
3020 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3021 before an outbound SMTP session. New log_selector, +tls_sni.
3023 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3024 NULL dereference. Report and patch from Alun Jones.
3026 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3027 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3028 Not seeing resolver debug output on NetBSD, but suspect this is a
3029 resolver implementation change.
3031 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3032 Left warnings. Added "eximon gdb" invocation mode.
3034 PP/21 Defaulting "accept_8bitmime" to true, not false.
3036 PP/22 Added -bw for inetd wait mode support.
3038 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3039 locate the relevant includes and libraries. Made this the default.
3041 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3042 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3044 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3045 This may cause build issues on older platforms.
3047 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3048 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3049 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3050 Added SNI support via GnuTLS too.
3051 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3053 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3055 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3056 Applied second patch from Janne, implementing suggestion to default
3057 multiple-strings-in-record handling to match SPF spec.
3059 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3061 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3062 read-only, out of scope).
3063 Patch from Wolfgang Breyha, report from Stuart Northfield.
3065 PP/29 Fix three issues highlighted by clang analyser static analysis.
3066 Only crash-plausible issue would require the Cambridge-specific
3067 iplookup router and a misconfiguration.
3068 Report from Marcin Mirosław.
3070 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3072 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3073 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3074 As part of this, removing so much warning spew let me fix some minor
3075 real issues in debug logging.
3077 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3078 assignment on my part. Fixed.
3080 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3081 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3082 Janne Snabb (who went above and beyond: thank you).
3084 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3085 string otherwise requires a connection and a bunch more work and it's
3086 relatively easy to get wrong. Should also expose TLS library linkage
3089 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3090 64-bit ${eval} (JH/03).
3092 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3093 GNU libc to support some of the 64-bit stuff, should not lead to
3094 conflicts. Defined before os.h is pulled in, so if a given platform
3095 needs to override this, it can.
3097 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3098 protection layer was required, which is not implemented.
3099 Bugzilla 1254, patch from Wolfgang Breyha.
3101 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3102 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3103 tls_dhparam take prime identifiers. Also unbreak combination of
3104 OpenSSL+DH_params+TLSSNI.
3106 PP/39 Disable SSLv2 by default in OpenSSL support.
3112 PP/01 Solaris build fix for Oracle's LDAP libraries.
3113 Bugzilla 1109, patch from Stephen Usher.
3115 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3117 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3120 TF/02 Fix a couple more cases where we did not log the error message
3121 when unlink() failed. See also change 4.74-TF/03.
3123 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3124 lock up or crash if it happened to be inside a call to libc when it
3125 got a SIGUSR1 from exiwhat.
3127 The SIGUSR1 handler appends the current process status to the process
3128 log which is later printed by exiwhat. It used to use the general
3129 purpose logging code to do this, but several functions it calls are
3130 not safe for signals.
3132 The new output code in the SIGUSR1 handler is specific to the process
3133 log, and simple enough that it's easy to inspect for signal safety.
3134 Removing some special cases also simplifies the general logging code.
3135 Removing the spurious timestamps from the process log simplifies
3138 TF/04 Improved ratelimit ACL condition.
3140 The /noupdate option has been deprecated in favour of /readonly which
3141 has clearer semantics. The /leaky, /strict, and /readonly update modes
3142 are mutually exclusive. The update mode is no longer included in the
3143 database key; it just determines when the database is updated. (This
3144 means that when you upgrade Exim will forget old rate measurements.)
3146 Exim now checks that the per_* options are used with an update mode that
3147 makes sense for the current ACL. For example, when Exim is processing a
3148 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3149 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3150 must specify per_mail/readonly. If you omit the update mode it defaults to
3151 /leaky where that makes sense (as before) or /readonly where required.
3153 The /noupdate option is now undocumented but still supported for
3154 backwards compatibility. It is equivalent to /readonly except that in
3155 ACLs where /readonly is required you may specify /leaky/noupdate or
3156 /strict/noupdate which are treated the same as /readonly.
3158 A useful new feature is the /count= option. This is a generalization
3159 of the per_byte option, so that you can measure the throughput of other
3160 aggregate values. For example, the per_byte option is now equivalent
3161 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3163 The per_rcpt option has been generalized using the /count= mechanism
3164 (though it's more complicated than the per_byte equivalence). When it is
3165 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3166 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3167 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3168 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3169 Note that using per_rcpt with a non-readonly update mode in more than
3170 one ACL will cause the recipients to be double-counted. (The per_mail
3171 and per_byte options don't have this problem.)
3173 The handling of very low rates has changed slightly. If the computed rate
3174 is less than the event's count (usually one) then this event is the first
3175 after a long gap. In this case the rate is set to the same as this event's
3176 count, so that the first message of a spam run is counted properly.
3178 The major new feature is a mechanism for counting the rate of unique
3179 events. The new per_addr option counts the number of different
3180 recipients that someone has sent messages to in the last time period. It
3181 behaves like per_rcpt if all the recipient addresses are different, but
3182 duplicate recipient addresses do not increase the measured rate. Like
3183 the /count= option this is a general mechanism, so the per_addr option
3184 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3185 example, measure the rate that a client uses different sender addresses
3186 with the options per_mail/unique=$sender_address. There are further
3187 details in the main documentation.
3189 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3191 TF/06 Removed a few PCRE remnants.
3193 TF/07 Automatically extract Exim's version number from tags in the git
3194 repository when doing development or release builds.
3196 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3197 Bugzilla 879. Patch from Paul Fisher.
3199 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3200 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3203 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3205 PP/05 Variable $av_failed, true if the AV scanner deferred.
3206 Bugzilla 1078. Patch from John Horne.
3208 PP/06 Stop make process more reliably on build failure.
3209 Bugzilla 1087. Patch from Heiko Schlittermann.
3211 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3212 Bugzilla 1089. Patch from Heiko Schlittermann.
3214 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3215 Bugzilla 1131. Patch from Holger Weiß.
3217 PP/09 Handle IPv6 addresses with SPF.
3218 Bugzilla 860. Patch from Wolfgang Breyha.
3220 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3222 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3225 PP/11 match_* no longer expand right-hand-side by default.
3226 New compile-time build option, EXPAND_LISTMATCH_RHS.
3227 New expansion conditions, "inlist", "inlisti".
3229 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3231 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3233 PP/14 fix log_write() format string regression from TF/03.
3234 Bugzilla 1152. Patch from Dmitry Isaikin.
3240 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3242 PP/02 Harmonised TLS library version reporting; only show if debugging.
3243 Layout now matches that introduced for other libraries in 4.74 PP/03.
3245 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3247 PP/04 New "dns_use_edns0" global option.
3249 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3252 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3253 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3255 TK/01 Updated PolarSSL code to 0.14.2.
3256 Bugzilla 1097. Patch from Andreas Metzler.
3258 PP/07 Catch divide-by-zero in ${eval:...}.
3259 Fixes bugzilla 1102.
3261 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3264 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3265 format-string attack -- SECURITY: remote arbitrary code execution.
3267 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3268 time unintentionally subject to list matching rules, letting the header
3269 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3270 arbitrary string expansion). This allowed for information disclosure.
3272 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3273 INT_MIN/-1 -- value coerced to INT_MAX.
3279 NM/01 Workaround for PCRE version dependency in version reporting
3282 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3283 This fixes portability to compilers other than gcc, notably
3284 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3286 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3287 makefiles for portability to HP-UX and POSIX correctness.
3289 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3290 Also via indented variable definition in the Makefile.
3291 (Debugging by Oliver Heesakkers).
3293 PP/02 Restore caching of spamd results with expanded spamd_address.
3294 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3296 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3297 Improves build reliability. Fix from: Frank Elsner
3299 NM/02 Fix wide character breakage in the rfc2047 coding
3300 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3302 NM/03 Allow underscore in dnslist lookups
3303 Fixes bug 1026. Patch from Graeme Fowler
3305 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3306 Code patches from Adam Ciarcinski of NetBSD.
3308 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3311 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3312 is logged, to avoid truncation. Patch from John Horne.
3314 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3315 Patch from Jakob Hirsch.
3317 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3318 SQL string expansion failure details.
3319 Patch from Andrey Oktyabrski.
3321 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3322 Patch from Simon Arlott.
3324 PP/09 New lookups functionality failed to compile on old gcc which rejects
3325 extern declarations in function scope.
3326 Patch from Oliver Fleischmann
3328 PP/10 Use sig_atomic_t for flags set from signal handlers.
3329 Check getgroups() return and improve debugging.
3330 Fixed developed for diagnosis in bug 927 (which turned out to be
3333 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3334 Patch from Mark Zealey.
3336 PP/12 Bugzilla 1056: Improved spamd server selection.
3337 Patch from Mark Zealey.
3339 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3340 Based on patch from Heiko Schlittermann.
3342 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3343 Patch from Uwe Doering, sign-off by Michael Haardt.
3345 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3346 variable declaration deep within a block. Bug and patch from
3349 PP/15 lookups-Makefile IRIX compatibility coercion.
3351 PP/16 Make DISABLE_DKIM build knob functional.
3353 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3354 Patch by Simon Arlott
3356 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3357 variable argument macros. Our copy now differs from upstream.
3363 TF/01 Failure to get a lock on a hints database can have serious
3364 consequences so log it to the panic log.
3366 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3367 controlled using the smtp_confirmation log selector.
3369 TF/03 Include the error message when we fail to unlink a spool file.
3371 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3372 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3373 for maintaining out-of-tree patches for some time.
3375 PP/01 Bugzilla 139: Documentation and portability issues.
3376 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3377 Handle per-OS dynamic-module compilation flags.
3379 PP/02 Let /dev/null have normal permissions.
3380 The 4.73 fixes were a little too stringent and complained about the
3381 permissions on /dev/null. Exempt it from some checks.
3382 Reported by Andreas M. Kirchwitz.
3384 PP/03 Report version information for many libraries, including
3385 Exim version information for dynamically loaded libraries. Created
3386 version.h, now support a version extension string for distributors
3387 who patch heavily. Dynamic module ABI change.
3389 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3390 privilege escalation vulnerability whereby the Exim run-time user
3391 can cause root to append content of the attacker's choosing to
3394 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3397 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3398 If dropping privileges for untrusted macros, we disabled normal logging
3399 on the basis that it would fail; for the Exim run-time user, this is not
3400 the case, and it resulted in successful deliveries going unlogged.
3401 Fixed. Reported by Andreas Metzler.
3407 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3408 only prepend for the Resent-* case. Fixes regression introduced in
3409 Exim 4.70 by NM/22 for Bugzilla 607.
3411 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3412 increasing numbers of administrators be bitten by this.
3414 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3416 PP/03 Bugzilla 994: added openssl_options main configuration option.
3418 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3420 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3422 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3424 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3425 without a peer certificate, leading to a segfault because of an
3426 assumption that peers always have certificates. Be a little more
3427 paranoid. Problem reported by Martin Tscholak.
3429 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3430 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3431 NB: ClamAV planning to remove STREAM in "middle of 2010".
3432 CL also introduces -bmalware, various -d+acl logging additions and
3433 more caution in buffer sizes.
3435 PP/09 Implemented reverse_ip expansion operator.
3437 PP/10 Bugzilla 937: provide a "debug" ACL control.
3439 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3441 PP/12 Bugzilla 973: Implement --version.
3443 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3445 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3447 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3449 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3450 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3451 ignore trailing whitespace.
3453 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3455 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3458 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3459 Notification from Dr Andrew Aitchison.
3461 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3462 ExtendedDetectionInfo response format.
3463 Notification from John Horne.
3465 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3468 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3469 XSL and documented dependency on system catalogs, with examples of how
3472 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3475 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3476 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3477 configuration file which is writeable by the Exim user or group.
3479 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3480 of configuration files to cover files specified with the -C option if
3481 they are going to be used with root privileges, not just the default
3484 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3485 option (effectively making it always true).
3487 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3488 files to be used while preserving root privileges.
3490 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3491 that rogue child processes cannot use them.
3493 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3494 run-time user, instead of root.
3496 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3497 Exim run-time user without dropping privileges.
3499 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3500 result string, instead of calling string_vformat() twice with the same
3503 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3504 for other users. Others should always drop root privileges if they use
3505 -C on the command line, even for a whitelisted configure file.
3507 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3509 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3515 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3516 $data_path, and $header_path variables; fixed documentation bugs and
3519 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3520 exipick to access non-standard spools, including the "frozen" queue
3523 NM/01 Bugzilla 965: Support mysql stored procedures.
3524 Patch from Alain Williams
3526 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3528 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3529 Patch from Andreas Metzler
3531 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3532 Patch from Kirill Miazine
3534 NM/05 Bugzilla 671: Added umask to procmail example.
3536 JJ/03 installed exipick 20100323.0, fixing doc bug
3538 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3539 directory. Notification and patch from Dan Rosenberg.
3541 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3543 TK/02 Improve log output when DKIM signing operation fails.
3545 MH/01 Treat the transport option dkim_domain as a colon separated
3546 list, not as a single string, and sign the message with each element,
3547 omitting multiple occurences of the same signer.
3549 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3550 Bugzilla 985, 986. Patch by Simon Arlott
3552 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3553 Patch by Simon Arlott
3555 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3556 MBX locking. Notification from Dan Rosenberg.
3562 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3564 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3566 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3568 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3570 NM/04 Bugzilla 915: Flag broken perl installation during build.
3576 TK/01 Added patch by Johannes Berg that expands the main option
3577 "spamd_address" if it starts with a dollar sign.
3579 TK/02 Write list of recipients to X-Envelope-Sender header when building
3580 the mbox-format spool file for content scanning (suggested by Jakob
3583 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3584 (http://www.dcc-servers.net/) support via dccifd. Activated by
3585 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3587 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3588 by Mark Daniel Reidel <mr@df.eu>.
3590 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3591 When building exim an external PCRE library is now needed -
3592 PCRE is a system library on the majority of modern systems.
3593 See entry on PCRE_LIBS in EDITME file.
3595 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3596 conversation. Added nologin parameter to request.
3597 Patch contributed by Kirill Miazine.
3599 TF/01 Do not log submission mode rewrites if they do not change the address.
3601 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3603 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3604 log files in place. Contributed by Roberto Lima.
3606 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3608 TF/03 Bugzilla 615: When checking the local_parts router precondition
3609 after a local_part_suffix or local_part_prefix option, Exim now
3610 does not use the address's named list lookup cache, since this
3611 contains cached lookups for the whole local part.
3613 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3614 Robert Millan. Documentation is in experimental-spec.txt.
3616 TF/04 Bugzilla 668: Fix parallel build (make -j).
3618 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3620 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3621 Patch provided by Jan Srzednicki.
3623 TF/05 Leading white space used to be stripped from $spam_report which
3624 wrecked the formatting. Now it is preserved.
3626 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3627 that they are available at delivery time.
3629 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3631 TF/08 TLS error reporting now respects the incoming_interface and
3632 incoming_port log selectors.
3634 TF/09 Produce a more useful error message if an SMTP transport's hosts
3635 setting expands to an empty string.
3637 NM/06 Bugzilla 744: EXPN did not work under TLS.
3638 Patch provided by Phil Pennock.
3640 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3641 Patch provided by Richard Godbee.
3643 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3644 acl_smtp_notquit, added index entry.
3646 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3647 Patch provided by Eugene Bujak.
3649 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3650 accept(). Patch provided by Maxim Dounin.
3652 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3653 Patch provided by Phil Pennock.
3655 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3657 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3658 Patch provided by Brad "anomie" Jorsch.
3660 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3661 Patch provided by Dean Brooks.
3663 TK/05 Add native DKIM support (does not depend on external libraries).
3665 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3666 Patch provided by Graeme Fowler.
3668 NM/16 Bugzilla 851: Documentation example syntax fix.
3670 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3672 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3675 NM/19 Bugzilla 745: TLS version reporting.
3676 Patch provided by Phil Pennock.
3678 NM/20 Bugzilla 167: bool: condition support.
3679 Patch provided by Phil Pennock.
3681 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3682 clients. Patch provided by Phil Pennock.
3684 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3685 Patch provided by Brad "anomie" Jorsch.
3687 NM/23 Bugzilla 687: Fix misparses in eximstats.
3688 Patch provided by Heiko Schlittermann.
3690 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3691 Patch provided by Heiko Schlittermann.
3693 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3694 plus update to original patch.
3696 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3698 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3699 Patch provided by David Brownlee.
3701 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3703 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3705 NM/30 Bugzilla 888: TLS documentation bugfixes.
3707 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3709 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3710 Unlike the original bugzilla I have changed all shell scripts in src tree.
3712 NM/33 Bugzilla 898: Transport filter timeout fix.
3713 Patch by Todd Rinaldo.
3715 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3716 Patch by Serge Demonchaux.
3718 NM/35 Bugzilla 39: Base64 decode bug fixes.
3719 Patch by Jakob Hirsch.
3721 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3723 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3725 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3727 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3733 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3734 ALT-N's libdkim that I have put here:
3735 http://duncanthrax.net/exim-experimental/
3737 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3738 were called 'true' and it seems that C99 defines that as a reserved
3739 keyword to be used with 'bool' variable types. That means you could
3740 not include C99-style headers which use bools without triggering
3741 build errors in sieve.c.
3743 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3744 as mailq or other aliases. Changed the --help handling significantly
3745 to do whats expected. exim_usage() emits usage/help information.
3747 SC/01 Added the -bylocaldomain option to eximstats.
3749 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3751 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3753 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3759 PH/01 Another patch from the Sieve maintainer.
3761 PH/02 When an IPv6 address is converted to a string for single-key lookup
3762 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3763 dots are used instead of colons so that keys in lsearch files need not
3764 contain colons. This was done some time before quoting was made available
3765 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3766 (notated using the quote facility) so as to distinguish them from IPv4
3767 keys. This meant that lookups for IP addresses in host lists did not work
3768 for iplsearch lookups.
3770 This has been fixed by arranging for IPv6 addresses to be expressed with
3771 colons if the lookup type is iplsearch. This is not incompatible, because
3772 previously such lookups could never work.
3774 The situation is now rather anomalous, since one *can* have colons in
3775 ordinary lsearch keys. However, making the change in all cases is
3776 incompatible and would probably break a number of configurations.
3778 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3781 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3782 conversion specification without a maximum field width, thereby enabling
3783 a rogue spamd server to cause a buffer overflow. While nobody in their
3784 right mind would setup Exim to query an untrusted spamd server, an
3785 attacker that gains access to a server running spamd could potentially
3786 exploit this vulnerability to run arbitrary code as the Exim user.
3788 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3789 $primary_hostname instead of what libspf2 thinks the hosts name is.
3791 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3792 a directory entry by the name of the lookup key. Previously, if a
3793 symlink pointed to a non-existing file or a file in a directory that
3794 Exim lacked permissions to read, a lookup for a key matching that
3795 symlink would fail. Now it is enough that a matching directory entry
3796 exists, symlink or not. (Bugzilla 503.)
3798 PH/03 The body_linecount and body_zerocount variables are now exported in the
3801 PH/04 Added the $dnslist_matched variable.
3803 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3804 This means they are set thereafter only if the connection becomes
3807 PH/06 Added the client_condition to authenticators so that some can be skipped
3808 by clients under certain conditions.
3810 PH/07 The error message for a badly-placed control=no_multiline_responses left
3811 "_responses" off the end of the name.
3813 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3815 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3816 (without spaces) instead of just copying the configuration text.
3818 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3820 PH/11 Added $max_received_linelength.
3822 PH/12 Added +ignore_defer and +include_defer to host lists.
3824 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3825 way in which PCRE > 7.0 is built.
3827 PH/14 Implemented queue_only_load_latch.
3829 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3830 MAIL command. The effect was to mangle the value on 64-bit systems.
3832 PH/16 Another patch from the Sieve maintainer.
3834 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3836 PH/18 If a system quota error occurred while trying to create the file for
3837 a maildir delivery, the message "Mailbox is full" was not appended to the
3838 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3839 only to a quota excession during the actual writing of the file.
3841 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3842 characters?) which causes problems in log lines. The DN values are now
3843 passed through string_printing() before being added to log lines.
3845 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3846 and InterBase are left for another time.)
3848 PH/21 Added message_body_newlines option.
3850 PH/22 Guard against possible overflow in moan_check_errorcopy().
3852 PH/23 POSIX allows open() to be a macro; guard against that.
3854 PH/24 If the recipient of an error message contained an @ in the local part
3855 (suitably quoted, of course), incorrect values were put in $domain and
3856 $local_part during the evaluation of errors_copy.
3862 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3863 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3866 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3867 issue a MAIL command.
3869 PH/02 In an ACL statement such as
3871 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3873 if a client was not listed at all, or was listed with a value other than
3874 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3875 the condition was not true (as it should be), so access was not denied.
3876 The bug was that the ! inversion was incorrectly passed on to the second
3877 item. This has been fixed.
3879 PH/03 Added additional dnslists conditions == and =& which are different from
3880 = and & when the dns lookup returns more than one IP address.
3882 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3883 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3885 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3886 FSYNC, which compiles an option called disable_fsync that allows for
3887 bypassing fsync(). The documentation is heavily laced with warnings.
3889 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3891 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3892 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3893 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3894 including adding "make clean"; (3) Added -fPIC when compiling the test
3895 dynamically loaded module, to get rid of a warning.
3897 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3898 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3899 = 0s. The bug is otherwise harmless.
3901 PH/07 There was a bug in the dovecot authenticator such that the value of
3902 $auth1 could be overwritten, and so not correctly preserved, after a
3903 successful authentication. This usually meant that the value preserved by
3904 the server_setid option was incorrect.
3906 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3908 PH/09 Installed PCRE release 7.0.
3910 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3911 run for batched SMTP input. It is now run at the start of every message
3912 in the batch. While fixing this I discovered that the process information
3913 (output by running exiwhat) was not always getting set for -bs and -bS
3914 input. This is fixed, and it now also says "batched" for BSMTP.
3916 PH/11 Added control=no_pipelining.
3918 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3919 patch, slightly modified), and move the expansion of helo_data till after
3920 the connection is made in the smtp transport (so it can use these
3923 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3925 PH/14 Added log_selector = +pid.
3927 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3929 PH/16 Add ${if forany and ${if forall.
3931 PH/17 Added dsn_from option to vary the From: line in DSNs.
3933 PH/18 Flush SMTP output before performing a callout, unless control =
3934 no_callout_flush is set.
3936 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3937 was true (the default) a successful delivery failed to delete the retry
3938 item, thus causing premature timeout of the address. The bug is now
3941 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3943 PH/21 Long custom messages for fakedefer and fakereject are now split up
3944 into multiline responses in the same way that messages for "deny" and
3945 other ACL rejections are.
3947 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3948 with slight modification.
3950 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3951 draft, changing the syntax and factoring some duplicate code".
3953 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3954 for deliveries of the second and subsequent messages over the same SMTP
3957 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3958 ${reduce, with only minor "tidies".
3960 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3962 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3963 expansion side effects.
3965 PH/27 When a message times out after an over-quota error from an Exim-imposed
3966 quota, the bounce message says "mailbox is full". This message was not
3967 being given when it was a system quota that was exceeded. It now should
3970 MH/03 Made $recipients available in local_scan(). local_scan() already has
3971 better access to the recipient list through recipients_list[], but
3972 $recipients can be useful in postmaster-provided expansion strings.
3974 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3975 in the case of a MAIL command with additional options following the
3976 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3977 were accidentally chopped off.
3979 PH/29 SMTP synchronization checks are implemented when a command is read -
3980 there is a check that no more input is waiting when there shouldn't be
3981 any. However, for some commands, a delay in an ACL can mean that it is
3982 some time before the response is written. In this time, more input might
3983 arrive, invalidly. So now there are extra checks after an ACL has run for
3984 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3985 pipelining has not been advertised.
3987 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3989 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3990 returned in the SMTP response when smtp_return_error_details was false.
3991 This has been fixed.
3993 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3994 instead of the C I/O that was originally supplied, because problems were
3995 reported on Solaris.
3997 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3998 Exim which did not show up earlier: it was assuming that a call to
3999 SSL_CTX_set_info_callback() might give an error value. In fact, there is
4000 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
4001 was a macro that became an assignment, so it seemed to work. This has
4002 changed to a proper function call with a void return, hence the compile
4003 error. Exim's code has been fixed.
4005 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4008 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4010 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4012 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4013 because it turns out to be too big or there is a timeout) there is no
4014 "Completed" line in the log. When some messages of this type were
4015 selected by exigrep, they were listed as "not completed". Others were
4016 picked up by some special patterns. I have improved the selection
4017 criteria to be more general.
4019 PH/38 The host_find_failed option in the manualroute router can now be set
4020 to "ignore", to completely ignore a host whose IP address cannot be
4021 found. If all hosts are ignored, the behaviour is controlled by the new
4022 host_all_ignored option.
4024 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4025 homing or because of multiple MX records with /mx) generated more than
4026 one IP address, and the following item turned out to be the local host,
4027 all the secondary addresses of the first item were incorrectly removed
4028 from the list, along with the local host and any following hosts (which
4029 is what is supposed to happen).
4031 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4032 whoever called Exim into the -H file. In the case of the daemon it was
4033 behaving confusingly. When first started, it used values for whoever
4034 started the daemon, but after a SIGHUP it used the Exim user (because it
4035 calls itself on a restart). I have changed the code so that it now always
4038 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4039 message are rejected with the same error (e.g. no authentication or bad
4040 sender address), and a DATA command is nevertheless sent (as can happen
4041 with PIPELINING or a stupid MUA), the error message that was given to the
4042 RCPT commands is included in the rejection of the DATA command. This is
4043 intended to be helpful for MUAs that show only the final error to their
4046 PH/42 Another patch from the Sieve maintainer.
4048 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4049 Eximstats - Fixed some broken HTML links and added missing column headers
4051 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4052 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4054 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4056 PH/43 Yet another patch from the Sieve maintainer.
4058 PH/44 I found a way to check for a TCP/IP connection going away before sending
4059 the response to the final '.' that terminates a message, but only in the
4060 case where the client has not sent further data following the '.'
4061 (unfortunately, this is allowed). However, in many cases there won't be
4062 any further data because there won't be any more messages to send. A call
4063 to select() can be used: if it shows that the input is "ready", there is
4064 either input waiting, or the socket has been closed. An attempt to read
4065 the next input character can distinguish the two cases. Previously, Exim
4066 would have sent an OK response which the client would never have see.
4067 This could lead to message repetition. This fix should cure that, at
4068 least in a lot of common cases.
4070 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4071 advertised in response to EHLO.
4077 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4078 fixed by 4.65/MH/01 (is this a record?) are fixed:
4080 (i) An empty string was always treated as zero by the numeric comparison
4081 operators. This behaviour has been restored.
4083 (ii) It is documented that the numeric comparison operators always treat
4084 their arguments as decimal numbers. This was broken in that numbers
4085 starting with 0 were being interpreted as octal.
4087 While fixing these problems I realized that there was another issue that
4088 hadn't been noticed. Values of message_size_limit (both the global option
4089 and the transport option) were treated as octal if they started with 0.
4090 The documentation was vague. These values are now always treated as
4091 decimal, and I will make that clear in the documentation.
4097 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4098 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4101 MH/01 Don't check that the operands of numeric comparison operators are
4102 integers when their expansion is in "skipping" mode (fixes bug
4103 introduced by 4.64-PH/07).
4105 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4106 child addresses, Exim now panics and dies. Previously, because the count
4107 is held in a short int, deliveries were likely to be lost. As such a
4108 large number of recipients for a single message is ridiculous
4109 (performance will be very, very poor), I have chosen to impose a limit
4110 rather than extend the field.
4116 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4117 leftover -K file (the existence of which was triggered by #402).
4118 While we were at it, introduced process PID as part of the -K
4119 filename. This should rule out race conditions when creating
4122 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4123 processing considerably. Previous code took too long for large mails,
4124 triggering a timeout which in turn triggers #401.
4126 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4127 in the DK code in transports.c. sendfile() is not really portable,
4128 hence the _LINUX specificness.
4130 TF/01 In the add_headers option to the mail command in an Exim filter,
4131 there was a bug that Exim would claim a syntax error in any
4132 header after the first one which had an odd number of characters
4135 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4136 callout verification, Exim cached a "reject" for the entire domain. This
4137 is correct for most verifications, but it is not correct for a recipient
4138 verification with use_sender or use_postmaster set, because in that case
4139 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4140 case of MAIL FROM:<> rejection from other early rejections (e.g.
4141 rejection of HELO). When verifying a recipient using a non-null MAIL
4142 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4143 Whatever the result of the callout, the value of the domain cache is
4144 left unchanged (for any other kind of callout, getting as far as trying
4145 RCPT means that the domain itself is ok).
4147 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4150 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4151 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4152 session, the daemon ignores ECONNECT errors and logs others; it now
4153 ignores EPIPE as well.
4155 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4156 (quoted-printable decoding).
4158 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4159 later the small subsequent patch to fix an introduced bug.
4161 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4163 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4165 PH/08 An error is now given if message_size_limit is specified negative.
4167 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4168 to be given (somewhat) arbitrary names.
4170 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4173 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4174 miscellaneous code fixes
4176 PH/10 Added the log_reject_target ACL modifier to specify where to log
4179 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4180 hostname. This is wrong, because it relates to the incoming message (and
4181 probably the interface on which it is arriving) and not to the outgoing
4182 callout (which could be using a different interface). This has been
4183 changed to use the value of the helo_data option from the smtp transport
4184 instead - this is what is used when a message is actually being sent. If
4185 there is no remote transport (possible with a router that sets up host
4186 addresses), $smtp_active_hostname is used.
4188 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4189 tweaks were necessary in order to get it to work (see also 21 below):
4190 (a) The code assumed that strncpy() returns a negative number on buffer
4191 overflow, which isn't the case. Replaced with Exim's string_format()
4193 (b) There were several signed/unsigned issues. I just did the minimum
4194 hacking in of casts. There is scope for a larger refactoring.
4195 (c) The code used strcasecmp() which is not a standard C function.
4196 Replaced with Exim's strcmpic() function.
4197 (d) The code set only $1; it now sets $auth1 as well.
4198 (e) A simple test gave the error "authentication client didn't specify
4199 service in request". It would seem that Dovecot has changed its
4200 interface. Fortunately there's a specification; I followed it and
4201 changed what the client sends and it appears to be working now.
4203 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4206 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4207 address A is aliased to B and C, where B exists and C does not. Without
4208 -v the output is "A verified" because verification stops after a
4209 successful redirection if more than one address is generated. However,
4210 with -v the child addresses are also verified. Exim was outputting "A
4211 failed to verify" and then showing the successful verification for C,
4212 with its parentage. It now outputs "B failed to verify", showing B's
4213 parentage before showing the successful verification of C.
4215 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4216 look up a TXT record in a specific list after matching in a combined
4219 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4220 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4221 they consult the DNS. I had assumed they would set it the way they
4222 wanted; and indeed my experiments on Linux seem to show that in some
4223 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4224 To be on the safe side, however, I have now made the interface to
4225 host_find_byname() similar to host_find_bydns(), with an argument
4226 containing the DNS resolver options. The host_find_byname() function now
4227 sets these options at its start, just as host_find_bydns() does. The smtp
4228 transport options dns_qualify_single and dns_search_parents are passed to
4229 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4230 of host_find_byname() use the default settings of RES_DEFNAMES
4231 (qualify_single) but not RES_DNSRCH (search_parents).
4233 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4234 spool_read_header() do less string testing, by means of a preliminary
4235 switch on the second character of optional "-foo" lines. (This is
4236 overdue, caused by the large number of possibilities that now exist.
4237 Originally there were few.) While I was there, I also converted the
4238 str(n)cmp tests so they don't re-test the leading "-" and the first
4239 character, in the hope this might squeeze out yet more improvement.
4241 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4242 flag allowing group syntax was set by the header_syntax check but not
4243 turned off, possible causing trouble later; (2) The flag was not being
4244 set at all for the header_verify test, causing "group"-style headers to
4245 be rejected. I have now set it in this case, and also caused header_
4246 verify to ignore an empty address taken from a group. While doing this, I
4247 came across some other cases where the code for allowing group syntax
4248 while scanning a header line wasn't quite right (mostly, not resetting
4249 the flag correctly in the right place). These bugs could have caused
4250 trouble for malformed header lines. I hope it is now all correct.
4252 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4253 with the "reply" argument non-NULL. The code, however (which originally
4254 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4255 but it didn't always do it. This confused somebody who was copying the
4256 code for some other use. I have removed all the tests.
4258 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4259 feature that was used to support insecure browsers during the U.S. crypto
4260 embargo. It requires special client support, and Exim is probably the
4261 only MTA that supported it -- and would never use it because real RSA is
4262 always available. This code has been removed, because it had the bad
4263 effect of slowing Exim down by computing (never used) parameters for the
4264 RSA_EXPORT functionality.
4266 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4267 authenticator to fail if there's a tab character in the incoming data
4268 (there should never be unless someone is messing about, as it's supposed
4269 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4270 if the connection is using TLS or if the remote IP is the same as the
4271 local IP, and the "valid-client-cert option" if a client certificate has
4274 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4275 authenticators. This can be used for authorization after authentication
4276 succeeds. (In the case of plaintext, it servers for both authentication
4279 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4280 if any retry times were supplied.
4282 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4283 connection, where there is no client IP address to check. In this
4284 situation, the verify now always succeeds.
4286 PH/25 Applied John Jetmore's -Mset patch.
4288 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4290 PH/27 In a string expansion for a processed (not raw) header when multiple
4291 headers of the same name were present, leading whitespace was being
4292 removed from all of them, but trailing whitespace was being removed only
4293 from the last one. Now trailing whitespace is removed from each header
4294 before concatenation. Completely empty headers in a concatenation (as
4295 before) are ignored.
4297 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4298 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4300 PH/29 [Removed. This was a change that I later backed out, and forgot to
4301 correct the ChangeLog entry (that I had efficiently created) before
4302 committing the later change.]
4304 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4305 address errors (4xx response to RCPT) over the same connection as other
4306 messages routed to the same hosts. Such deliveries are always "forced",
4307 so retry times are not inspected. This resulted in far too many retries
4308 for the affected addresses. The effect occurred only when there were more
4309 hosts than the hosts_max_try setting in the smtp transport when it had
4310 the 4xx errors. Those hosts that it had tried were not added to the list
4311 of hosts for which the message was waiting, so if all were tried, there
4312 was no problem. Two fixes have been applied:
4314 (i) If there are any address or message errors in an SMTP delivery, none
4315 of the hosts (tried or untried) are now added to the list of hosts
4316 for which the message is waiting, so the message should not be a
4317 candidate for sending over the same connection that was used for a
4318 successful delivery of some other message. This seems entirely
4319 reasonable: after all the message is NOT "waiting for some host".
4320 This is so "obvious" that I'm not sure why it wasn't done
4321 previously. Hope I haven't missed anything, but it can't do any
4322 harm, as the worst effect is to miss an optimization.
4324 (ii) If, despite (i), such a delivery is accidentally attempted, the
4325 routing retry time is respected, so at least it doesn't keep
4326 hammering the server.
4328 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4329 in ${readsocket because some servers need this prod.
4331 PH/32 Added some extra debug output when updating a wait-xxx database.
4333 PH/33 The hint "could be header name not terminated by colon", which has been
4334 given for certain expansion errors for a long time, was not being given
4335 for the ${if def:h_colon_omitted{... case.
4337 PH/34 The spec says: "With one important exception, whenever a domain list is
4338 being scanned, $domain contains the subject domain." There was at least
4339 one case where this was not true.
4341 PH/35 The error "getsockname() failed: connection reset by peer" was being
4342 written to the panic log as well as the main log, but it isn't really
4343 panic-worthy as it just means the connection died rather early on. I have
4344 removed the panic log writing for the ECONNRESET error when getsockname()
4347 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4348 runs only) independently of the message's sender address. This meant
4349 that, if the 4xx error was in fact related to the sender, a different
4350 message to the same recipient with a different sender could confuse
4351 things. In particular, this can happen when sending to a greylisting
4352 server, but other circumstances could also provoke similar problems.
4353 I have changed the default so that the retry time for these errors is now
4354 based a combination of the sender and recipient addresses. This change
4355 can be overridden by setting address_retry_include_sender=false in the
4358 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4359 remote server are returned as part of bounce messages. This was not
4360 happening for LMTP over a pipe (the lmtp transport), but now it is the
4361 same for both kinds of LMTP.
4363 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4364 in header lines that were in fact CNAMEs. This is no longer the case.
4366 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4367 and queue runs started by the daemon processed all messages. This has
4368 been fixed so that -R and -S can now usefully be given with -q<time>.
4370 PH/40 Import PCRE release 6.7 (fixes some bugs).
4372 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4374 PH/42 Give an error if -q is specified more than once.
4376 PH/43 Renamed the variables $interface_address and $interface_port as
4377 $received_ip_address and $received_port, to make it clear that these
4378 values apply to message reception, and not to the outgoing interface when
4379 a message is delivered. (The old names remain recognized, of course.)
4381 PH/44 There was no timeout on the connect() call when using a Unix domain
4382 socket in the ${readsocket expansion. There now is.
4384 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4385 be meaningful with "accept".
4387 SC/01 Eximstats V1.43
4388 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4390 SC/02 Eximstats V1.44
4391 Use a glob alias rather than an array ref in the generated
4392 parser. This improves both readability and performance.
4394 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4395 Collect SpamAssassin and rejection statistics.
4396 Don't display local sender or destination tables unless
4397 there is data to show.
4398 Added average volumes into the top table text output.
4400 SC/04 Eximstats V1.46
4401 Collect data on the number of addresses (recipients)
4402 as well as the number of messages.
4404 SC/05 Eximstats V1.47
4405 Added 'Message too big' to the list of mail rejection
4406 reasons (thanks to Marco Gaiarin).
4408 SC/06 Eximstats V1.48
4409 Mainlog lines which have GMT offsets and are too short to
4410 have a flag are now skipped.
4412 SC/07 Eximstats V1.49 (Alain Williams)
4413 Added the -emptyok flag.
4415 SC/08 Eximstats V1.50
4416 Fixes for obtaining the IP address from reject messages.
4418 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4419 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4420 whitespace changes from 4.64-PH/27
4422 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4425 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4426 are found, allow negative numbers in numeric criteria)
4428 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4430 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4431 to match changes made in 4.64-PH/43
4433 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4435 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4436 "domain" is actually an IP address, and force a failure. This locks out
4437 those revolvers/nameservers that support "A-for-A" lookups, in
4438 contravention of the specifications.
4440 PH/48 When a host name was looked up from an IP address, and the subsequent
4441 forward lookup of the name timed out, the host name was left in
4442 $sender_host_name, contrary to the specification.
4444 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4445 restricted to single-key lookups, Exim was not diagnosing an error if
4446 * or *@ was used with a query-style lookup.
4448 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4450 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4451 long ago, but noone interested enough thought of it. Let's just say that
4452 the "1.1" means that there are some new functions that weren't there at
4453 some point in the past.
4455 PH/51 Error processing for expansion failure of helo_data from an smtp
4456 transport during callout processing was broken.
4458 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4459 tested/used via the -bh/-bhc/-bs options.
4461 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4462 bug, fixed in subsequent PCRE releases).
4464 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4465 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4467 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4473 SC/01 Use a glob alias rather than an array ref in eximstats generated
4474 parser. This improves both readability and performance.
4476 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4477 Don't display local sender or destination tables in eximstats unless
4478 there is data to show.
4479 Added average volumes into the eximstats top table text output.
4481 SC/03 Collect data on the number of addresses (recipients) as well
4482 as the number of messages in eximstats.
4484 TF/01 Correct an error in the documentation for the redirect router. Exim
4485 does (usually) call initgroups() when daemonizing.
4487 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4488 with consistent privilege compared to when running as a daemon.
4490 TF/03 Note in the spec that $authenticated_id is not set for local
4491 submissions from trusted users.
4493 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4494 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4496 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4497 by adding some example configuration directives to the default
4498 configuration file. A little bit of work is required to uncomment the
4499 directives and define how usernames and passwords are checked, but
4500 there is now a framework to start from.
4502 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4503 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4504 without this. I don't know how relevant this is to other LDAP libraries.
4506 PH/02 Add the verb name to the "unknown ACL verb" error.
4508 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4510 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4512 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4513 directory not expanded when it should be if an expanded home directory
4514 was set for the address (which is overridden by the transport).
4516 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4519 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4520 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4521 because it is too late at that time, and has no effect.
4523 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4524 security issue with \' (bugzilla #107). I could not use the
4525 PQescapeStringConn() function, because it needs a PGconn value as one of
4528 PH/08 When testing addresses using -bt, indicate those final addresses that
4529 are duplicates that would not cause an additional delivery. At least one
4530 person was confused, thinking that -bt output corresponded to deliveries.
4531 (Suppressing duplicates isn't a good idea as you lose the information
4532 about possibly different redirections that led to the duplicates.)
4534 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4535 systems where poll() doesn't work, in particular OS X.
4537 PH/10 Added more information to debugging output for retry time not reached.
4539 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4540 operations in malware.c.
4542 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4545 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4546 syslog_duplication set false, log lines that would normally be written
4547 both the the main log and to the reject log were not written to syslog at
4550 PH/14 In the default configuration, change the use of "message" in ACL warn
4551 statements to "add_header".
4553 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4554 not followed by a command (e.g. "seen endif").
4556 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4557 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4560 PH/17 Added extra conditions to the default value of delay_warning_condition
4564 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4565 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4566 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4569 The Auto-Submitted: and various List- headers are standardised, whereas I
4570 don't think Precedence: ever was.
4572 PH/18 Refactored debugging code in route_finduser() to show more information,
4573 in particular, the error code if getpwnam() issues one.
4575 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4576 This is apparently needed in addition to the PH/07 change above to avoid
4577 any possible encoding problems.
4579 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4580 but not after initializing Perl.
4582 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4583 output them only if debugging. By default they are written stderr,
4584 apparently, which is not desirable.
4586 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4589 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4592 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4594 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4595 authenticated or an ident call has been made. Suppress the default
4596 values for $authenticated_id and $authenticated_sender (but permit -oMai
4597 and -oMas) when testing with -bh.
4599 PH/24 Re-jigged the order of the tests in the default configuration so that the
4600 tests for valid domains and recipients precede the DNS black list and CSA
4601 tests, on the grounds that those ones are more expensive.
4603 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4604 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4605 This bug exists in every version of Exim that I still have, right back to
4608 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4609 However, an attempt to turn on case-sensitivity in a regex key by
4610 including (?-i) didn't work because the subject string was already
4611 lowercased, and the effects were non-intuitive. It turns out that a
4612 one-line patch can be used to allow (?-i) to work as expected.
4618 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4619 other effects) broke the use of negated acl sub-conditions.
4621 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4624 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4625 "Deny" causes Exim to reject the incoming connection with a 554 error.
4626 Unfortunately, if there is a major crisis, such as a disk failure,
4627 tcp-wrappers gives "deny", whereas what one would like would be some
4628 kind of temporary error. A kludge has been added to help with this.
4629 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4630 554 error is used if errno is still zero or contains ENOENT (which occurs
4631 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4634 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4636 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4637 errors. Otherwise a message that provokes a temporary error (when other
4638 messages do not) can cause a whole host to time out.
4640 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4641 addresses were routed directly to files or pipes from a redirect router.
4642 File deliveries just didn't batch; pipe deliveries might have suffered
4645 PH/06 A failure to get a lock for a hints database would erroneously always say
4646 "Failed to get write lock", even when it was really a read lock.
4648 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4649 of 0600. This has been changed to use the value of the lockfile_mode
4650 option (which defaults to 0600).
4652 PH/08 Applied small patch from the Sieve maintainer.
4654 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4655 folder from quota calculations, a direct delivery into this folder messed
4656 up the contents of the maildirsize file. This was because the regex was
4657 used only to exclude .Trash (or whatever) when the size of the mailbox
4658 was calculated. There was no check that a delivery was happening into an
4659 excluded directory. This bug has been fixed by ignoring all quota
4660 processing for deliveries into excluded directories.
4662 PH/10 Added the maildirfolder_create_regex option to appendfile.
4668 PH/01 The code for finding all the local interface addresses on a FreeBSD
4669 system running IPv6 was broken. This may well have applied to all BSD
4670 systems, as well as to others that have similar system calls. The broken
4671 code found IPv4 interfaces correctly, but gave incorrect values for the
4672 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4673 that it would not match correctly against @[] and not recognize the IPv6
4676 PH/02 The ipliteral router was not recognizing addresses of the form user@
4677 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4679 PH/03 Added disable_ipv6.
4681 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4682 decoded form, because it is most often used to construct To: headers
4683 lines in autoreplies, and the decoded form may well be syntactically
4684 invalid. However, $reply_address has leading white space removed, and all
4685 newlines turned into spaces so that the autoreply transport does not
4688 PH/05 If group was specified without a user on a router, and no group or user
4689 was specified on a transport, the group from the router was ignored.
4691 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4692 for visible compile-time settings that can be used to change these
4693 numbers, for those that want even more. Backwards compatibility with old
4694 spool files has been maintained. However, going back to a previous Exim
4695 release will lost any variables that are in spool files.
4697 PH/07 Two small changes when running in the test harness: increase delay when
4698 passing a TCP/IP connection to a new process, in case the original
4699 process has to generate a bounce, and remove special handling of
4700 127.0.0.2 (sic), which is no longer necessary.
4702 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4703 be the same on different OS.
4705 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4708 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4709 whether --show-vars was specified or not
4711 JJ/02 exipick: Added support for new ACL variable spool format introduced
4714 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4715 syntactically invalid From: or Reply-to: line, and a filter used this to
4716 generate an autoreply, and therefore failed to obtain an address for the
4717 autoreply, Exim could try to deliver to a non-existent relative file
4718 name, causing unrelated and misleading errors. What now happens is that
4719 it logs this as a hard delivery error, but does not attempt to create a
4722 PH/11 The exinext utility has a -C option for testing purposes, but although
4723 the given file was scanned by exinext itself; it wasn't being passed on
4724 when Exim was called.
4726 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4727 an end-of-file indication when reading a command response.
4729 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4730 compiled. In many other places in Exim, IPv6 addresses are always
4731 recognized, so I have changed this. It also means that IPv4 domain
4732 literals of the form [IPV4:n.n.n.n] are now always recognized.
4734 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4735 used if the router is not running as root, for example, when verifying at
4736 ACL time, or when using -bh. The debugging output from this situation was
4737 non-existent - all you got was a failure to exec. I have made two
4740 (a) Failures to set uid/gid, the current directory, or a process leader
4741 in a subprocess such as that created by queryprogram now generate
4742 suitable debugging output when -d is set.
4744 (b) The queryprogram router detects when it is not running as root,
4745 outputs suitable debugging information if -d is set, and then runs
4746 the subprocess without attempting to change uid/gid.
4748 PH/15 Minor change to Makefile for building test_host (undocumented testing
4751 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4752 additional section of a DNS packet that returns MX or SRV records.
4753 Instead, it always explicitly searches for A/AAAA records. This avoids
4754 major problems that occur when a DNS server includes only records of one
4755 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4756 fixed another bug: if SRV records were looked up and the corresponding
4757 address records were *not* found in the additional section, the port
4758 values from the SRV records were lost.
4760 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4761 using the correct key (the original address) when searching the retry
4762 rules in order to find which one to use for generating the retry hint.
4764 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4765 adding the default one. Similarly, if it contains a Reply-To: header, the
4766 errors_reply_to option, if set, is not used.
4768 PH/19 When calculating a retry time, Exim used to measure the "time since
4769 failure" by looking at the "first failed" field in the retry record. Now
4770 it does not use this if it is later than than the arrival time of the
4771 message. Instead it uses the arrival time. This makes for better
4772 behaviour in cases where some deliveries succeed, thus re-setting the
4773 "first failed" field. An example is a quota failure for a huge message
4774 when small messages continue to be delivered. Without this change, the
4775 "time since failure" will always be short, possible causing more frequent
4776 delivery attempts for the huge message than are intended.
4777 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4779 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4780 $1, $2, $3) because the numerical variables can be reset during some
4781 expansion items (e.g. "match"), thereby losing the authentication data.
4783 PH/21 Make -bV show the size of off_t variables so that the test suite can
4784 decide whether to run tests for quotas > 2G.
4786 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4787 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4788 filecount value is greater than 2G or if a quota value is greater than 2G
4789 on a system where the size of off_t is not greater than 4, a panic error
4792 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4793 never match. The debug and -bh output now contains an explicit error
4794 message indicating a malformed IPv4 address or mask.
4796 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4797 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4798 PH/23 above applies.
4800 PH/25 Do not write to syslog when running in the test harness. The only
4801 occasion when this arises is a failure to open the main or panic logs
4802 (for which there is an explicit test).
4804 PH/26 Added the /no_tell option to "control=freeze".
4806 PH/27 If a host name lookup failed very early in a connection, for example, if
4807 the IP address matched host_lookup and the reverse lookup yielded a name
4808 that did not have a forward lookup, an error message of the form "no IP
4809 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4810 could be logged. Now it outputs the IP address instead of "NULL".
4812 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4813 allows the sender and the authenticated sender to be set when
4814 submitting a message from within Exim. Since child_open_exim() is
4815 documented for local_scan(), the new function should be too.
4817 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4818 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4819 results in an empty string is now treated as unset.
4821 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4823 PH/31 Added sender_verify_fail logging option.
4825 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4826 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4827 tidied the source and removed it altogether.
4829 PH/33 When a queue run was abandoned because the load average was too high, a
4830 log line was always written; now it is written only if the queue_run log
4831 selector is set. In addition, the log line for abandonment now contains
4832 information about the queue run such as the pid. This is always present
4833 in "start" and "stop" lines but was omitted from the "abandon" line.
4835 PH/34 Omit spaces between a header name and the colon in the error message that
4836 is given when verify = headers_syntax fails (if there are lots of them,
4837 the message gets confusing).
4839 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4840 names, as there are now some PTR records that contain slashes. This check
4841 is only to protect against broken name servers that fall over on strange
4842 characters, so the fact that it applies to all lookups doesn't matter.
4844 PH/36 Now that the new test suite is complete, we can remove some of the
4845 special code in Exim that was needed for the old test suite. For example,
4846 sorting DNS records because real resolvers return them in an arbitrary
4847 order. The new test suite's fake resolver always returns records in the
4850 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4851 bounces) except when queue_only is set, to avoid logging races between
4852 the different processes.
4854 PH/38 Panic-die if .include specifies a non-absolute path.
4856 PH/39 A tweak to the "H" retry rule from its user.
4858 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4859 a label. They prevented compilation on older perls.
4861 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4862 a warning to be raised on newish perls.
4864 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4865 on queue. Changes to match documented behaviour of showing count of
4866 messages matching specified criteria.
4868 PH/40 Changed the default ident timeout from 30s to 5s.
4870 PH/41 Added support for the use of login_cap features, on those BSD systems
4871 that have them, for controlling the resources used by pipe deliveries.
4873 PH/42 The content-scanning code uses fopen() to create files in which to put
4874 message data. Previously it was not paying any attention to the mode of
4875 the files. Exim runs with umask(0) because the rest of the code creates
4876 files with open(), and sets the required mode explicitly. Thus, these
4877 files were ending up world-writeable. This was not a big issue, because,
4878 being within the spool directory, they were not world-accessible. I have
4879 created a function called modefopen, which takes an additional mode
4880 argument. It sets umask(777), creates the file, chmods it to the required
4881 mode, then resets the umask. All the relevant calls to fopen() in the
4882 content scanning code have been changed to use this function.
4884 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4885 to 24 hours. This avoids potential overflow problems when processing G
4886 and H retry rules. I suspect nobody ever tinkers with this value.
4888 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4890 PH/45 When the plaintext authenticator is running as a client, the server's
4891 challenges are checked to ensure they are valid base64 strings. By
4892 default, the authentication attempt is cancelled if an invalid string is
4893 received. Setting client_ignore_invalid_base64 true ignores these errors.
4894 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4895 they are received. Thus, the responses can be made to depend on the
4896 challenges. If an invalid string is ignored, an empty string is placed in
4899 PH/46 Messages that are created by the autoreply transport now contains a
4900 References: header, in accordance with RFCs 2822 and 3834.
4902 PH/47 Added authenticated_sender_force to the smtp transport.
4904 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4906 PH/49 Installed latest patch from the Sieve maintainer.
4908 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4909 was also set, the appendfile transport was unnecessarily scanning a
4910 directory of message files (e.g. for maildir delivery) to find the count
4911 of files (along with the size), even though it did not need this
4912 information. It now does the scan only if it needs to find either the
4913 size of the count of files.
4915 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4917 PH/52 Two bugs concerned with error handling when the smtp transport is
4920 (i) Exim was not creating retry information for temporary errors given
4921 for individual recipients after the DATA command when the smtp transport
4922 was used in LMTP mode. This meant that they could be retried too
4923 frequently, and not timed out correctly.
4925 (ii) Exim was setting the flag that allows error details to be returned
4926 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4927 recipients that were returned after the DATA command.
4929 PH/53 This is related to PH/52, but is more general: for any failing address,
4930 when detailed error information was permitted to be returned to the
4931 sender, but the error was temporary, then after the final timeout, only
4932 "retry timeout exceeded" was returned. Now it returns the full error as
4933 well as "retry timeout exceeded".
4935 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4936 do this, and (what is worse) MTAs that accept it.
4938 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4939 will now be deprecated.
4941 PH/56 New os.c-cygwin from the Cygwin maintainer.
4943 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4944 formats (previously only available in exim formats via -bpr, -bpru,
4945 and -bpra. Now also available in native and exiqgrep formats)
4947 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4948 with very large, slow to parse queues
4950 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4952 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4954 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4955 responses to authentication challenges, though it was showing the
4956 challenges; (ii) I've removed the CR characters from the debug output for
4959 PH/58 Allow for the insertion of a newline as well as a space when a string
4960 is turned into more than one encoded-word during RFC 2047 encoding. The
4961 Sieve code now uses this.
4963 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4964 data_4xx, lost_connection, tls_required.
4966 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4967 message was being sent as an SMTP response.
4969 PH/61 Add -l and -k options to exicyclog.
4971 PH/62 When verifying, if an address was redirected to one new address, so that
4972 verification continued, and the new address failed or deferred after
4973 having set something in $address_data, the value of $address_data was not
4974 passed back to the ACL. This was different to the case when no
4975 redirection occurred. The value is now passed back in both cases.
4977 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4978 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4979 use login_cap.h, so on its own it isn't the distinguishing feature. The
4980 new name refers directly to the setclassresources() function.
4982 PH/65 Added configuration files for NetBSD3.
4984 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4986 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4987 is preferred over IPv4.
4989 PH/68 The bounce_return_message and bounce_return_body options were not being
4990 honoured for bounces generated during the reception of non-SMTP messages.
4991 In particular, this applied to messages rejected by the ACL. This bug has
4992 been fixed. However, if bounce_return_message is true and bounce_return_
4993 body is false, the headers that are returned for a non-SMTP message
4994 include only those that have been read before the error was detected.
4995 (In the case of an ACL rejection, they have all been read.)
4997 PH/69 The HTML version of the specification is now built in a directory called
4998 spec_html instead of spec.html, because the latter looks like a path with
4999 a MIME-type, and this confuses some software.
5001 PH/70 Catch two compiler warnings in sieve.c.
5003 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5004 function verify_get_ident() calls ip_connect() to connect a socket, but
5005 if the "connect()" function timed out, ip_connect() used to close the
5006 socket. However, verify_get_ident() also closes the socket later, and in
5007 between Exim writes to the log, which may get opened at this point. When
5008 the socket was closed in ip_connect(), the log could get the same file
5009 descriptor number as the socket. This naturally causes chaos. The fix is
5010 not to close the socket in ip_connect(); the socket should be closed by
5011 the function that creates it. There was only one place in the code where
5012 this was missing, in the iplookup router, which I don't think anybody now
5013 uses, but I've fixed it anyway.
5015 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5016 well as to direct DNS lookups. Otherwise the handling of names in host
5017 lists is inconsistent and therefore confusing.
5023 PH/01 Two changes to the default runtime configuration:
5025 (1) Move the checks for relay_from_hosts and authenticated clients from
5026 after to before the (commented out) DNS black list checks.
5028 (2) Add control=submission to the relay_from_hosts and authenticated
5029 clients checks, on the grounds that messages accepted by these
5030 statements are most likely to be submissions.
5032 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5034 (1) Generate an error if the third argument for the ${prvs expansion is
5037 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5040 (3) Reset the variables that are obtained from the first argument of
5041 ${prvscheck and used in the second argument before leaving the code,
5042 because their memory is reclaimed, so using them afterwards may do
5045 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5046 one (it's much easier than Tom thought :-).
5048 (5) Because of (4), we can now allow for the use of $prvscheck_result
5049 inside the third argument.
5051 PH/03 For some reason, the default setting of PATH when running a command from
5052 a pipe transport was just "/usr/bin". I have changed it to
5055 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5056 anything to be listed in the output from -bV.
5058 PH/05 When a filter generated an autoreply, the entire To: header line was
5059 quoted in the delivery log line, like this:
5061 => >A.N.Other <ano@some.domain> <original@ddress> ...
5063 This has been changed so that it extracts the operative address. There
5064 may be more than one such address. If so, they are comma-separated, like
5067 => >ano@some.domain,ona@other.domain <original@ddress> ...
5069 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5070 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5071 not being looked up in the rDNS to get a host name, Exim was showing the
5072 IP address twice in Received: lines, even though the IP addresses were
5073 identical. For example:
5075 Received: from [1.2.3.4] (helo=[1.2.3.4])
5077 However, if the real host name was known, it was omitting the HELO data
5078 if it matched the actual IP address. This has been tidied up so that it
5079 doesn't show the same IP address twice.
5081 PH/07 When both +timestamp and +memory debugging was on, the value given by
5082 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5083 called by the debug printing, thereby overwriting the timestamp buffer.
5084 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5086 PH/08 When the original message was included in an autoreply transport, it
5087 always said "this is a copy of the message, including all the headers",
5088 even if body_only or headers_only was set. It now gives an appropriate
5091 PH/09 Applied a patch from the Sieve maintainer which:
5093 o fixes some comments
5094 o adds the (disabled) notify extension core
5095 o adds some debug output for the result of if/elsif tests
5096 o points to the current vacation draft in the documentation
5097 and documents the missing references header update
5101 o fixes a bug in processing the envelope test (when testing
5102 multiple envelope elements, the last element determined the
5105 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5106 Electronic Mail") by including:
5108 Auto-submitted: auto-generated
5110 in the messages that it generates (bounce messages and others, such as
5111 warnings). In the case of bounce messages for non-SMTP messages, there was
5112 also a typo: it was using "Auto_submitted" (underscore instead of
5113 hyphen). Since every message generated by Exim is necessarily in response
5114 to another message, thes have all been changed to:
5116 Auto-Submitted: auto-replied
5118 in accordance with these statements in the RFC:
5120 The auto-replied keyword:
5122 - SHOULD be used on messages sent in direct response to another
5123 message by an automatic process,
5125 - MUST NOT be used on manually-generated messages,
5127 - MAY be used on Delivery Status Notifications (DSNs) and Message
5128 Disposition Notifications (MDNs),
5130 - MUST NOT be used on messages generated by automatic or periodic
5131 processes, except for messages which are automatic responses to
5134 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5135 to the default Received: header definition.
5137 PH/12 Added log selector acl_warn_skipped (default on).
5139 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5140 variables because (a) they are in the wrong storage pool and (b) even if
5141 they were copied, it wouldn't work properly because of the caching.
5143 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5144 checking when decoding. Apparently there are clients that generate
5145 overlong encoded strings. Why am I not surprised?
5147 PH/15 If the first argument of "${if match_address" was not empty, but did not
5148 contain an "@" character, Exim crashed. Now it writes a panic log message
5149 and treats the condition as false.
5151 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5153 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5154 extension "envelope-auth". The code is finished and in agreement with
5155 other implementations, but there is no documentation so far and in fact,
5156 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5157 not changing the active code.
5159 Print executed "if" and "elsif" statements when debugging is used. This
5160 helps a great deal to understand what a filter does.
5162 Document more things not specified clearly in RFC3028. I had all this
5163 sorted out, when out of a sudden new issues came to my mind. Oops."
5165 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5168 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5169 canonical form (as documented). However, after a host name lookup from
5170 the IP address, check_host() was doing a simple string comparison with
5171 addresses acquired from the DNS when checking that the found name did
5172 have the original IP as one of its addresses. Since any found IPv6
5173 addresses are likely to be in abbreviated form, the comparison could
5174 fail. Luckily, there already exists a function for doing the comparison
5175 by converting both addresses to binary, so now that is used instead of
5176 the text comparison.
5178 PH/20 There was another similar case to PH/19, when a complete host name was
5179 given in a host list; looking up its IP address could give an abbreviated
5180 form, whereas the current host's name might or might not be abbreviated.
5181 The same fix has been applied.
5187 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5188 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5191 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5192 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5194 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5196 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5197 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5198 Also, for "Auto-Submitted", treat anything other than "no" as
5199 non-personal, in accordance with RFC 3834. (Previously it treated
5200 anything starting "auto-" as non-personal.)
5202 TF/01 The control=submission/name=... option had a problem with syntax
5203 errors if the name included a slash character. The /name= option
5204 now slurps the rest of the string, so it can include any characters
5205 but it must come last in the list of options (after /sender_retain
5208 PH/05 Some modifications to the interface to the fake nameserver for the new
5216 TK/01 Added the "success_on_redirect" address verification option. See
5217 NewStuff for rationale and an example.
5219 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5221 PH/02 Patch to exigrep to allow it to work on syslog lines.
5223 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5224 fread() to skip over the body file's header line, because in Cygwin the
5225 header line is locked and is inaccessible.
5227 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5228 co-exist for some time) to make it clear that it is the Exim ID that is
5229 referenced, not the Message-ID: header line.
5231 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5232 string_format() function, because snprintf() does not exist on all
5235 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5236 ${stat: expansion item.
5238 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5239 protocol synchronization error", to keep the pedants happy.
5241 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5242 well as for IRIX systems, when gcc is being used. See the host.c source
5245 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5247 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5250 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5251 transport and to the smtp transport in LMTP mode.
5253 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5255 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5256 run a filter in a subprocess. This could lead to confusion in subsequent
5257 lookups in the parent process. There should also be a search_tidyup() at
5258 the end of the subprocess.
5260 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5261 only if the host matched helo_try_verify_hosts, which caused the
5262 verification to occur when the EHLO/HELO command was issued. The ACL just
5263 tested the remembered result. Now, if a previous verification attempt has
5264 not happened, "verify = helo" does it there and then.
5266 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5268 TK/03 Fix log output including CR from clamd.
5270 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5271 exist provoked a memory error which could cause a segfault.
5273 PH/15 Installed PCRE 6.2
5275 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5277 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5278 of the problem. Specifically, suggested +O2 rather than +O1 for the
5281 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5283 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5284 record, the port was not correctly logged when the outgoing_port log
5285 selector was set (it logged the transort's default port).
5287 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5288 fallback_hosts, and "hosts" in the smtp transport.
5290 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5291 host errors such as "Connection refused".
5293 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5294 authentication with radiusclient 0.4.9:
5296 - Error returned from rc_read_config was caught wrongly
5297 - Username/password not passed on to radius server due to wrong length.
5299 The presumption is that some radiusclient API changes for 4.51/PH/17
5300 were not taken care of correctly. The code is still untested by me (my
5301 Linux distribution still has 0.3.2 of radiusclient), but it was
5302 contributed by a Radius user.
5304 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5305 expanding the "port" option of the smtp transport.
5307 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5308 while reading a MIME header. Thanks to Tom Hughes for a patch.
5310 PH/24 Include config.h inside local_scan.h so that configuration settings are
5313 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5314 that in an ACL for RCPT (for example), you can examine exactly what was
5317 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5318 commands, but it was not correctly comparing the address with the actual
5319 client host address. Thus, it would show the EHLO address in Received:
5320 header lines when this was not necessary.
5322 PH/27 Added the % operator to ${eval:}.
5324 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5325 it should be ignoring failures (because with -C, for example, it has lost
5326 privilege). It wasn't ignoring creation failures other than "already
5329 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5332 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5333 because an input line was too long, either on its own, or by virtue of
5334 too many continuations, the temporary file was not being removed, and the
5335 return code was incorrect.
5337 PH/31 Missing "BOOL" in function definition in filtertest.c.
5339 PH/32 Applied Sieve patches from the maintainer.
5341 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5343 PH/33 Added "verify = not_blind".
5345 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5346 Local/Makefile (with some defaults set). These are used in built scripts
5347 such as exicyclog, but they have never been used in the exim_install
5348 script (though there are many overriding facilities there). I have
5349 arranged that the exim_install script now takes note of these two
5352 PH/35 Installed configuration files for Dragonfly.
5354 PH/36 When a locally submitted message by a trusted user did not contain a
5355 From: header, and the sender address was obtained from -f or from an SMTP
5356 MAIL command, and the trusted user did not use -F to supply a sender
5357 name, $originator_name was incorrectly used when constructing a From:
5358 header. Furthermore, $originator_name was used for submission mode
5359 messages from external hosts without From: headers in a similar way,
5360 which is clearly wrong.
5362 PH/37 Added control=suppress_local_fixups.
5364 PH/38 When log_selector = +received_sender was set, and the addition of the
5365 sender made the log line's construction buffer exactly full, or one byte
5366 less than full, an overflow happened when the terminating "\n" was
5369 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5370 when the result of a list match is failure because a DNS lookup failed.
5372 PH/40 RM_COMMAND is now used in the building process.
5374 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5375 the "build-* directories that it finds.
5377 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5378 address in a domain literal was a prefix of an interface address.
5380 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5381 when verifying a sender address, unless rewrite_headers is false.
5383 PH/44 Wrote a long comment about why errors_to addresses are verified as
5384 recipients, not senders.
5386 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5387 the ratelimit ACL was added.
5389 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5391 PH/46 Added extra information about PostgreSQL errors to the error string.
5393 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5394 avoiding the need to install special zones in a real server. This is
5395 backwards compatible; if it can't find the fake resolver, it drops back.
5396 Thus, both old and new test suites can be run.
5398 TF/02 Added util/ratelimit.pl
5400 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5401 clock is set back in time.
5403 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5404 Candler <B.Candler@pobox.com>.
5406 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5407 OK for addresses that are the result of redirections.
5409 PH/48 A number of further additions for the benefit of the new test suite,
5410 including a fake gethostbyname() that interfaces to the fake DNS resolver
5413 TF/06 The fix for widen_domains has also been applied to qualify_single and
5414 search_parents which are the other dnslookup options that can cause
5417 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5420 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5422 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5423 service" if it encountered an MX record with an empty target hostname.
5424 The message is now "an MX or SRV record indicated no SMTP service".
5426 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5427 if the DNS of the sending site is misconfigured. This is quite a
5428 common situation. This change restores the behaviour of treating a
5429 helo verification defer as a failure.
5431 PH/51 If self=fail was set on a router, the bounce message did not include the
5432 actual error message.
5438 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5440 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5441 command itself ended in error, the underlying message about the transport
5442 filter timeout was being overwritten with the pipe command error. Now the
5443 underlying error message should be appended to the second error message.
5445 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5447 PH/02 Reduce the amount of output that "make" produces by default. Full output
5448 can still be requested.
5450 PH/03 The warning log line about a condition test deferring for a "warn" verb
5451 was being output only once per connection, rather than after each
5452 occurrence (because it was using the same function as for successful
5453 "warn" verbs). This seems wrong, so I have changed it.
5455 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5456 it should not have, which might have caused a crash in the right
5457 circumstances, but probably never did.
5459 PH/04 Installed a modified version of Tony Finch's patch to make submission
5460 mode fix the return path as well as the Sender: header line, and to
5461 add a /name= option so that you can make the user's friendly name appear
5464 TF/03 Added the control = fakedefer ACL modifier.
5466 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5467 Mark Lowes for thorough testing.
5469 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5471 TK/03 Merged latest SRS patch from Miles Wilton.
5473 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5474 with the definition in sysexits.h (which is #included earlier).
5475 Fortunately, Exim does not actually use EX_OK. The code used to try to
5476 preserve the sysexits.h value, by assuming that macro definitions were
5477 scanned for macro replacements. I have been disabused of this notion,
5478 so now the code just undefines EX_OK before #including unistd.h.
5480 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5481 in the smtp transport. When a block could not be written in a single
5482 write() function, the timeout was being re-applied to each part-write.
5483 This seems wrong - if the receiver was accepting one byte at a time it
5484 would take for ever. The timeout is now adjusted when this happens. It
5485 doesn't have to be particularly precise.
5487 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5488 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5490 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5491 without a domain if the check to <postmaster@domain> fails.
5493 SC/01 Eximstats: added -xls and the ability to specify output files
5494 (patch written by Frank Heydlauf).
5496 SC/02 Eximstats: use FileHandles for outputting results.
5498 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5500 SC/04 Eximstats: fixed display of large numbers with -nvr option
5502 SC/05 Eximstats: fixed merging of reports with empty tables.
5504 SC/06 Eximstats: added the -include_original_destination flag
5506 SC/07 Eximstats: removed tabs and trailing whitespace.
5508 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5510 TK/06 MBOX spool code: Add real "From " MBOX separator line
5511 so the .eml file is really in mbox format (even though
5512 most programs do not really care). Patch from Alex Miller.
5514 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5515 The latter is generated from $received_to and is only set if the
5516 message has one envelope recipient. SA can use these headers,
5517 obviously out-of-the-box. Patch from Alex Miller.
5519 PH/08 The ${def test on a variable was returning false if the variable's
5520 value was "0", contrary to what the specification has always said!
5521 The result should be true unless the variable is empty.
5523 PH/09 The syntax error of a character other than { following "${if
5524 def:variable_name" (after optional whitespace) was not being diagnosed.
5525 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5526 accidental colon was present, for example, could give incorrect results.
5528 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5529 result is used (not including appendfile, where other changes are about
5532 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5533 This involved changing a lot of size variables from int to off_t. It
5534 should work with maildirs and everything.
5536 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5537 spamd dying while we are connected to it.
5539 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5542 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5543 function for BDB 4.3.
5545 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5547 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5548 the sources for PCRE have been re-arranged and more files are now
5551 PH/15 The code I had for printing potentially long long variables in PH/11
5552 above was not the best (it lost precision). The length of off_t variables
5553 is now inspected at build time, and an appropriate printing format (%ld
5554 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5555 to be "long long int" or "long int". This is needed for the internal
5556 formatting function string_vformat().
5558 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5559 the configuration file to be ":syslog", then the script "guesses" where
5560 the logs files are, rather than using the compiled in default. In our
5561 case the guess is not the same as the compiled default, so the script
5562 suddenly stopped working when I started to use syslog. The patch checks
5563 to see if log_file_path is "". If so, it attempts to read it from exim
5564 with no configuration file to get the compiled in version, before it
5565 falls back to the previous guessing code."
5567 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5568 implementing BATV in an Exim configuration. See NewStuff for the gory
5571 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5572 Makefile that are specific to HP-UX.
5574 PH/18 If the "use_postmaster" option was set for a recipient callout together
5575 with the "random" option, the postmaster address was used as the MAIL
5576 FROM address for the random test, but not for the subsequent recipient
5577 test. It is now used for both.
5579 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5580 patch removes a few documentation additions to RFC 3028, because the
5581 latest draft now contains them. It adds the new en;ascii-case comparator
5582 and a new error check for 8bit text in MIME parts. Comparator and
5583 require names are now matched exactly. I enabled the subaddress
5584 extension, but it is not well tested yet (read: it works for me)."
5586 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5587 rework some of the code of TK/09 above to avoid the hardwired use of
5588 "%lld" and "long long". Replaced the call to snprintf() with a call to
5591 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5592 records point to non-existent hosts", "retry timeout exceeded", and
5593 "retry time not reached for any host after a long failure period".
5595 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5596 experimental DomainKeys support:
5598 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5599 (2) On an error such as an illegally used "control", the wrong name for
5600 the control was given.
5602 These problems did NOT occur unless DomainKeys support was compiled.
5604 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5606 PH/24 Added ${if match_ip condition.
5608 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5609 databases so that it will be absolutely obvious if a crash occurs in the
5610 DB library. This is a regular occurrence (often caused by mis-matched
5613 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5614 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5615 picked up on a user's system that detects such things. There doesn't seem
5616 to be a gcc warning option for this - only an attribute that has to be
5617 put on the function's prototype. It seems that in Fedora Core 4 they have
5618 set this on a number of new functions. No doubt there will be more in due
5621 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5622 specify a transport. However, if an address that was verified by such a
5623 router was the subject of a callout, Exim crashed because it tried to
5624 read the rcpt_include_affixes from the non-existent transport. Now it
5625 just assumes that the setting of that option is false. This bug was
5626 introduced by 4.51/PH/31.
5628 PH/28 Changed -d+all to exclude +memory, because that information is very
5629 rarely of interest, but it makes the output a lot bigger. People tend to
5630 do -d+all out of habit.
5632 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5633 code in os-type was giving problems when libc.so lives in lib64, like on
5636 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5637 aren't the modern standard, and it seems that some systems' include files
5638 don't always have them. Exim was already checking for some of the newer
5639 ones like T_AAAA, and defining it itself. I've added checks for all the
5640 record types that Exim uses.
5642 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5643 not automatically generating a new one, as it is supposed to. This
5644 prevented TLS from working. If the file did exist, but contained invalid
5645 data, a new version was generated, as expected. It was only the case of a
5646 non-existent file that was broken.
5648 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5649 with a change in libdomainkeys > 0.64.
5651 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5652 from DNS. If the selector record carries the flag, it now has
5653 precedence over the domain-wide flag.
5655 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5657 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5658 the use of a port name that isn't defined in /etc/services) occurred, the
5659 message was deferred as in a normal delivery, and thus remained on the
5660 spool, instead of being failed because of the mua_wrapper setting. This
5661 is now fixed, and I tidied up some of the mua_wrapper messages at the
5664 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5665 the messages in a hash of arrays rather than using individual hashes.
5666 This is a bit cleaner and results in dramatic memory savings, albeit
5667 at a slight CPU cost.
5669 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5670 as requested by Marc Sherman.
5672 SC/10 Eximstats: added histograms for user specified patterns as requested
5675 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5677 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5678 fopen() in the content-scanning modules that did not already have it.
5684 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5685 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5687 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5689 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5691 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5692 to test Sieve filters that use "vacation".
5694 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5695 that changes the way the GnuTLS parameters are stored in the cache file.
5696 The new format can be generated externally. For backward compatibility,
5697 if the data in the cache doesn't make sense, Exim assumes it has read an
5698 old-format file, and it generates new data and writes a new file. This
5699 means that you can't go back to an older release without removing the
5702 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5703 work if there are any delivery delays because "one_time" forces the
5704 parent to be marked "delivered", so its unseen clone is never tried
5705 again. For this reason, Exim now forbids the simultaneous setting of
5708 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5709 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5710 entry if you want to know the details. The fix, however, neglected to
5711 consider the case when local delivery batching is involved. The test for
5712 "previously delivered" was not happening when checking to see if an
5713 address could be batched with a previous (undelivered) one; under
5714 certain circumstances this could lead to multiple deliveries to the same
5717 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5718 in its include files, and this causes problems building Exim.
5720 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5721 header_syntax) but Exim was just ignoring anything given after a slash.
5722 In particular, this caused confusion with an attempt to use "verify =
5723 reverse_host_lookup/defer_ok". An error is now given when options are
5724 supplied for verify items that do not have them. (Maybe reverse_host_
5725 lookup should have a defer_ok option, but that's a different point.)
5727 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5728 defined by RFC 821) to 2048, because there were problems with some AUTH
5729 commands, and RFC 1869 says the size should be increased for extended
5730 SMTP commands that take arguments.
5732 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5735 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5736 "unknown" error; now it says that the functionality isn't in the binary.
5738 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5739 an address' error message when a string expansion fails (syntax or
5740 whatever). Otherwise the password may appear in the log. Following change
5741 PH/42 below, there is no longer a chance of it appearing in a bounce
5744 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5746 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5747 of its addresses was ever tried. (Bugzilla bug #2.)
5749 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5750 the result incorrectly in the debug output. (It correctly added a newline
5751 to what was transported.)
5753 TF/01 Added $received_time.
5755 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5756 commented out examples of how to interface to a virus scanner and to
5757 SpamAssassin. Also added commented examples of av_scanner and
5758 spamd_address settings.
5760 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5761 and controls are allowed in which ACLs. There were a couple of minor
5762 errors. Some of the entries in the conditions table (which is a table of
5763 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5764 negation of where the condition IS allowed.
5766 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5768 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5769 header file does not have a version number, so I've had to invent a new
5770 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5771 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5772 radiusclient), but it was contributed by a Radius user.
5774 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5775 files or directories, for OpenSSL.
5777 PH/19 When an Exim process that is running as root has to create an Exim log
5778 file, it does so in a subprocess that runs as exim:exim so as to get the
5779 ownership right at creation (otherwise, other Exim processes might see
5780 the file with the wrong ownership). There was no test for failure of this
5781 fork() call, which would lead to the process getting stuck as it waited
5782 for a non-existent subprocess. Forks do occasionally fail when resources
5783 run out. I reviewed all the other calls to fork(); they all seem to check
5786 PH/20 When checking for unexpected SMTP input at connect time (before writing
5787 the banner), Exim was not dealing correctly with a non-positive return
5788 from the read() function. If the client had disconnected by this time,
5789 the result was a log entry for a synchronization error with an empty
5790 string after "input=" when read() returned zero. If read() returned -1
5791 (an event I could not check), uninitialized data bytes were printed.
5792 There were reports of junk text (parts of files, etc) appearing after
5795 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5797 PH/22 Added support for macro redefinition, and (re)definition in between
5798 driver and ACL definitions.
5800 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5801 forgetting to use the resulting value; it was using the unexpanded value.
5803 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5804 hadn't been configured. The fix is from Juergen Kreileder, who
5805 understands it better than I do:
5807 "Here's what I see happening with three configured cyrus_sasl
5808 authenticators configured (plain, login, cram-md5):
5810 On startup auth_cyrus_sasl_init() gets called for each of these.
5811 This means three calls to sasl_listmech() without a specified mech_list.
5812 => SASL tests which mechs of all available mechs actually work
5813 => three warnings about OTP not working
5814 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5816 With the patch, sasl_listmech() also gets called three times. But now
5817 SASL's mech_list option is set to the server_mech specified in the the
5818 authenticator. Or in other words, the answer from sasl_listmech()
5819 gets limited to just the mech you're testing for (which is different
5821 => the return list contains just 'plain' or 'login', 'cram-md5' or
5822 nothing depending on the value of ob->server_mech.
5824 I've just tested the patch: Authentication still works fine,
5825 unavailable mechs specified in the exim configuration are still
5826 caught, and the auth.log warnings about OTP are gone."
5828 PH/25 When debugging is enabled, the contents of the command line are added
5829 to the debugging output, even when log_selector=+arguments is not
5832 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5833 answer is "GNU", and only if the return is "GNU/something" is the answer
5836 PH/27 $acl_verify_message is now set immediately after the failure of a
5837 verification in an ACL, and so is available in subsequent modifiers. In
5838 particular, the message can be preserved by coding like this:
5840 warn !verify = sender
5841 set acl_m0 = $acl_verify_message
5843 Previously, $acl_verify_message was set only while expanding "message"
5844 and "log_message" when a very denied access.
5846 PH/28 Modified OS/os.c-Linux with
5848 -#ifndef OS_LOAD_AVERAGE
5849 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5851 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5852 nomenclature these days.)
5854 PH/29 Installed patch from the Sieve maintainer that adds the options
5855 sieve_useraddress and sieve_subaddress to the redirect router.
5857 PH/30 In these circumstances:
5858 . Two addresses routed to the same list of hosts;
5859 . First host does not offer TLS;
5860 . First host accepts first address;
5861 . First host gives temporary error to second address;
5862 . Second host offers TLS and a TLS session is established;
5863 . Second host accepts second address.
5864 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5865 and peerdn, if requested) that were in fact used only for the second
5868 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5869 attention to any local part prefix or suffix that was matched by the
5870 router that accepted the address. It now behaves in the same way as it
5871 does for delivery: the affixes are removed from the local part unless
5872 rcpt_include_affixes is set on the transport.
5874 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5875 timeout during the DATA phase of an incoming message.
5877 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5878 applied a patch sanctioned by the Sieve maintainer.
5880 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5881 the uid or gid is negative. A case of a negative gid caused this to be
5882 noticed. The fix allows for either to be negative.
5884 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5885 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5888 PH/36 The change PH/12 above was broken. Fixed it.
5890 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5891 the grounds that routing the same address twice would always produce the
5892 same answer. This might have been true once, but it is certainly no
5893 longer true now. Routing a child address may depend on the previous
5894 routing that produced that child. Some complicated redirection strategies
5895 went wrong when messages had multiple recipients, and made Exim's
5896 behaviour dependent on the order in which the addresses were given.
5898 I have moved the duplicate checking until after the routing is complete.
5899 Exim scans the addresses that are assigned to local and remote
5900 transports, and removes any duplicates. This means that more work will be
5901 done, as duplicates will always all be routed, but duplicates are
5902 presumably rare, so I don't expect this is of any significance.
5904 For deliveries to pipes, files, and autoreplies, the duplicate checking
5905 still happens during the routing process, since they are not going to be
5908 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5909 It corrects a timeout issue with spamd. This is Ian's comment: "The
5910 background is that sometimes spamd either never reads data from a
5911 connection it has accepted, or it never writes response data. The exiscan
5912 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5913 blindly assumes that writes won't block so it may never time out."
5915 PH/39 Allow G after quota size as well as K and M.
5917 PH/40 The value set for $authenticated_id in an authenticator may not contain
5918 binary zeroes or newlines because the value is written to log lines and
5919 to spool files. There was no check on this. Now the value is run through
5920 the string_printing() function so that such characters are converted to
5921 printable escape sequences.
5923 PH/41 $message_linecount is a new variable that contains the total number of
5924 lines in the message. Compare $body_linecount, which is the count for the
5927 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5928 bounce and delay warning messages, except in certain special cases, which
5931 (a) An SMTP error message from a remote host;
5932 (b) A message specified in a :fail: redirection;
5933 (c) A message specified in a "fail" command in a system filter;
5934 (d) A message specified in a FAIL return from the queryprogram router;
5935 (e) A message specified by the cannot_route_message router option.
5937 In these cases only, Exim does include the error details in bounce and
5938 warning messages. There are also a few cases where bland messages such
5939 as "unrouteable address" or "local delivery error" are given.
5941 PH/43 $value is now also set for the "else" part of a ${run expansion.
5943 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5944 being worked on, but at least Exim now implements the latest version to
5947 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5948 process to complete was treated as a delivery failure, a timeout while
5949 writing the message to the pipe was logged, but erroneously treated as a
5950 successful delivery. Such timeouts include transport filter timeouts. For
5951 consistency with the overall process timeout, these timeouts are now
5952 treated as errors, giving rise to delivery failures by default. However,
5953 there is now a new Boolean option for the pipe transport called
5954 timeout_defer, which, if set TRUE, converts the failures into defers for
5955 both kinds of timeout. A transport filter timeout is now identified in
5958 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5959 systems where "make" and "gmake" are different, calling "gmake" at top
5960 level broke things. I've arranged for the value of $(MAKE) to be passed
5961 from the Makefile to this script so that it can call the same version of
5965 A note about Exim versions 4.44 and 4.50
5966 ----------------------------------------
5968 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5969 changes of various kinds. As a consequence, a big documentation update was
5970 needed. This delayed the release for rather longer than seemed good, especially
5971 in the light of a couple of (minor) security issues. Therefore, the changes
5972 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5973 release. So 4.44 and 4.50 are in effect two different branches that both start
5976 I have left the 4.50 change log unchanged; it contains all the changes since
5977 4.43. The change log for 4.44 is below; many of its items are identical to
5978 those for 4.50. This seems to be the most sensible way to preserve the
5979 historical information.
5985 1. Minor wording change to the doc/README.SIEVE file.
5987 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5988 computation of the current number of files was incorrect.
5990 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5991 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5994 4. Give more explanation in the error message when the command for a transport
5995 filter fails to execute.
5997 5. There are several places where Exim runs a non-Exim command in a
5998 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5999 was being done only for the command run by the queryprogram router. It is
6000 now done for all such subprocesses. The other cases are: ${run, transport
6001 filters, and the commands run by the lmtp and pipe transports.
6003 6. Added CONFIGURE_GROUP build-time option.
6005 7. Some older OS have a limit of 256 on the maximum number of file
6006 descriptors. Exim was using setrlimit() to set 1000 as a large value
6007 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6008 systems. I've change it so that if it can't get 1000, it tries for 256.
6010 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6011 was an oversight, and furthermore, ever since the addition of extra
6012 controls (e.g. 4.43/32), the checks on when to allow different forms of
6013 "control" were broken. There should now be diagnostics for all cases when a
6014 control that does not make sense is encountered.
6016 9. Added the /retain_sender option to "control=submission".
6018 10. $recipients is now available in the predata ACL (oversight).
6020 11. Tidy the search cache before the fork to do a delivery from a message
6021 received from the command line. Otherwise the child will trigger a lookup
6022 failure and thereby defer the delivery if it tries to use (for example) a
6023 cached ldap connection that the parent has called unbind on.
6025 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6026 of $address_data from the recipient verification was clobbered by the
6027 sender verification.
6029 13. The value of address_data from a sender verification is now available in
6030 $sender_address_data in subsequent conditions in the ACL statement.
6032 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6034 15. Added a new option "connect=<time>" to callout options, to set a different
6037 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6038 was its contents. (It was OK if the option was not defined at all.)
6040 17. A "Completed" log line is now written for messages that are removed from
6041 the spool by the -Mrm option.
6043 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6044 information about exactly what failed.
6046 19. Added -dd to debug only the daemon process.
6048 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6049 handles timeouts, both on the server side and network timeouts. Renamed the
6050 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6052 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6053 It is now set to "smtps".
6055 22. $host_address is now set to the target address during the checking of
6056 ignore_target_hosts.
6058 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6059 being passed; this would have caused $sender_host_name to have been used if
6060 matching the list had actually called for a host name (not very likely,
6061 since this list is usually IP addresses). A host name is now passed as
6064 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6065 code that specifies a non-restarting handler (typically sigaction() in
6066 modern systems) in an attempt to fix a rare and obscure crash bug.
6068 25. Narrowed the window for a race in the daemon that could cause it to ignore
6069 SIGCHLD signals. This is not a major problem, because they are used only to
6070 wake it up if nothing else does.
6072 26. A malformed maildirsize file could cause Exim to calculate negative values
6073 for the mailbox size or file count. Odd effects could occur as a result.
6074 The maildirsize information is now recalculated if the size or filecount
6077 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6078 support for a long time. Removed HAVE_SYS_VFS_H.
6080 28. Installed the latest version of exipick from John Jetmore.
6082 29. In an address list, if the pattern was not a regular expression, an empty
6083 subject address (from a bounce message) matched only if the pattern was an
6084 empty string. Non-empty patterns were not even tested. This was the wrong
6085 because it is perfectly reasonable to use an empty address as part of a
6086 database query. An empty address is now tested by patterns that are
6087 lookups. However, all the other forms of pattern expect the subject to
6088 contain a local part and a domain, and therefore, for them, an empty
6089 address still always fails if the pattern is not itself empty.
6091 30. Exim went into a mad DNS loop when attempting to do a callout where the
6092 host was specified on an smtp transport, and looking it up yielded more
6093 than one IP address.
6095 31. Re-factored the code for checking spool and log partition space into a
6096 function that finds that data and another that does the check. The former
6097 is then used to implement four new variables: $spool_space, $log_space,
6098 $spool_inodes, and $log_inodes.
6100 32. The RFC2047 encoding function was originally intended for short strings
6101 such as real names; it was not keeping to the 75-character limit for
6102 encoded words that the RFC imposes. It now respects the limit, and
6103 generates multiple encoded words if necessary. To be on the safe side, I
6104 have increased the buffer size for the ${rfc2047: expansion operator from
6107 33. It is now permitted to omit both strings after an "if" condition; if the
6108 condition is true, the result is "true". As before, when the second string
6109 is omitted, a false condition yields an empty string. This makes it less
6110 cumbersome to write custom ACL and router conditions.
6112 34. Failure to deliver a bounce message always caused it to be frozen, even if
6113 there was an errors_to setting on the router. The errors_to setting is now
6116 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6117 canonical form (fully expanded) before being placed in
6118 $sender_host_address.
6120 36. The table in the code that translates DNS record types into text (T_A to
6121 "A" for instance) was missing entries for NS and CNAME. It is just possible
6122 that this could have caused confusion if both these types were looked up
6123 for the same domain, because the text type is used as part of Exim's
6124 per-process caching. But the chance of anyone hitting this buglet seems
6127 37. The dnsdb lookup has been extended in a number of ways.
6129 (1) There is a new type, "zns", which walks up the domain tree until it
6130 finds some nameserver records. It should be used with care.
6132 (2) There is a new type, "mxh", which is like "mx" except that it returns
6133 just the host names, not the priorities.
6135 (3) It is now possible to give a list of domains (or IP addresses) to be
6136 looked up. The behaviour when one of the lookups defers can be
6137 controlled by a keyword.
6139 (4) It is now possible to specify the separator character for use when
6140 multiple records are returned.
6142 38. The dnslists ACL condition has been extended: it is now possible to supply
6143 a list of IP addresses and/or domains to be looked up in a particular DNS
6146 39. Added log_selector=+queue_time_overall.
6148 40. When running the queue in the test harness, wait just a tad after forking a
6149 delivery process, to get repeatability of debugging output.
6151 41. Include certificate and key file names in error message when GnuTLS fails
6152 to set them up, because the GnuTLS error message doesn't include the name
6153 of the failing file when there is a problem reading it.
6155 42. Allow both -bf and -bF in the same test run.
6157 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6159 44. The "Exiscan patch" is now merged into the mainline Exim source.
6161 45. Sometimes the final signoff response after QUIT could fail to get
6162 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6163 < 0 before doing a fflush(). This bug looks as though it goes back to the
6164 introduction of TLS in release 3.20, but "sometimes" must have been rare
6165 because the tests only now provoked it.
6167 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6168 (this can affect the format of dates).
6170 47. exim_tidydb, when checking for the continued existence of a message for
6171 which it has found a message-specific retry record, was not finding
6172 messages that were in split spool directories. Consequently, it was
6173 deleting retry records that should have stayed in existence.
6175 48. Steve fixed some bugs in eximstats.
6177 49. The SPA authentication driver was not abandoning authentication and moving
6178 on to the next authenticator when an expansion was forced to fail,
6179 contradicting the general specification for all authenticators. Instead it
6180 was generating a temporary error. It now behaves as specified.
6182 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6183 (the order specifies the preference for clients). The order is now AES256,
6184 AES128, 3DES, ARCFOUR128.
6186 51. Small patch to Sieve code - explicitly set From: when generating an
6189 52. Exim crashed if a remote delivery caused a very long error message to be
6190 recorded - for instance if somebody sent an entire SpamAssassin report back
6191 as a large number of 550 error lines. This bug was coincidentally fixed by
6192 increasing the size of one of Exim's internal buffers (big_buffer) that
6193 happened as part of the Exiscan merge. However, to be on the safe side, I
6194 have made the code more robust (and fixed the comments that describe what
6197 53. Now that there can be additional text after "Completed" in log lines (if
6198 the queue_time_overall log selector is set), a one-byte patch to exigrep
6199 was needed to allow it to recognize "Completed" as not the last thing in
6202 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6203 patch that reportedly fixes this has been added. I am not expert enough to
6204 create a test for it. This is what the patch creator wrote:
6206 "I found a little strange behaviour of ldap code when working with
6207 Windows 2003 AD Domain, where users was placed in more than one
6208 Organization Units. When I tried to give exim partial DN, the exit code
6209 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6210 But simultaneously result of request was absolutely normal ldap result,
6211 so I produce this patch..."
6213 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6214 REFERENCE, so I have modified the code to exclude the patch when that macro
6217 55. Some experimental protocols are using DNS PTR records for new purposes. The
6218 keys for these records are domain names, not reversed IP addresses. The
6219 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6220 leaves it alone. Component reversal etc. now happens only for IP addresses.
6223 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6225 57. Double the size of the debug message buffer (to 2048) so that more of very
6226 long debug lines gets shown.
6228 58. The exicyclog utility now does better if the number of log files to keep
6229 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6231 59. Two changes related to the smtp_active_hostname option:
6233 (1) $smtp_active_hostname is now available as a variable.
6234 (2) The default for smtp_banner uses $smtp_active_hostname instead
6235 of $primary_hostname.
6237 60. The host_aton() function is supposed to be passed a string that is known
6238 to be a valid IP address. However, in the case of IPv6 addresses, it was
6239 not checking this. This is a hostage to fortune. Exim now panics and dies
6240 if the condition is not met. A case was found where this could be provoked
6241 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6242 components; fortuitously, this particular loophole had already been fixed
6243 by change 4.50/55 above.
6245 If there are any other similar loopholes, the new check in host_aton()
6246 itself should stop them being exploited. The report I received stated that
6247 data on the command line could provoke the exploit when Exim was running as
6248 exim, but did not say which command line option was involved. All I could
6249 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6250 running as the user.
6253 61. There was a buffer overflow vulnerability in the SPA authentication code
6254 (which came originally from the Samba project). I have added a test to the
6255 spa_base64_to_bits() function which I hope fixes it.
6258 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6259 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6261 63. The daemon start-up calls getloadavg() while still root for those OS that
6262 need the first call to be done as root, but it missed one case: when
6263 deliver_queue_load_max is set with deliver_drop_privilege. This is
6264 necessary for the benefit of the queue runner, because there is no re-exec
6265 when deliver_drop_privilege is set.
6267 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6268 This has been fixed.
6270 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6271 was in use, was not putting the data itself into the right store pool;
6272 consequently, it could be overwritten for a subsequent message in the same
6273 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6276 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6278 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6279 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6280 the function were treating the return as a boolean value, which happened to
6281 work because 0=false and not-0=true, but is not correct code.
6283 68. The host_aton() function was not handling scoped IPv6 addresses (those
6284 with, for example, "%eth0" on the end) correctly.
6286 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6287 negated items (that is, ~something) in unsigned ints. Some compilers
6288 apparently mutter when there is no cast.
6290 70. If an address verification called from an ACL failed, and did not produce a
6291 user-specific message (i.e. there was only a "system" message), nothing was
6292 put in $acl_verify_message. In this situation, it now puts the system
6295 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6296 session; change 4.31/43 added the unwanted input to the log line - except
6297 that it did not do this in the start of session case. It now does.
6299 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6300 This is wrong and can cause the other end to generate a synchronization
6301 error if it is another Exim or anything else that does the synchronization
6302 check. A QUIT command is no longer sent after a timeout.
6304 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6305 during host lookups.
6307 74. The defer_ok option of callout verification was not working if it was used
6308 when verifying addresses in header lines, that is, for this case:
6310 verify = header_sender/callout=defer_ok
6312 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6313 those file descriptors could be used for SMTP connections. If anything
6314 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6315 could be sent to the SMTP client, causing chaos. The daemon now opens
6316 stdin, stdout, and stderr to /dev/null when it puts itself into the
6319 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6320 log by default. The user can override this with suitable Perl magic.
6322 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6323 the log message when discard triggers, was not working for the DATA ACL or
6324 for the non-SMTP ACL.
6326 78. Error message wording change in sieve.c.
6328 79. If smtp_accept_max_per_host was set, the number of connections could be
6329 restricted to fewer than expected, because the daemon was trying to set up
6330 a new connection before checking whether the processes handling previous
6331 connections had finished. The check for completed processes is now done
6332 earlier. On busy systems, this bug wouldn't be noticed because something
6333 else would have woken the daemon, and it would have reaped the completed
6336 80. If a message was submitted locally by a user whose login name contained one
6337 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6338 It caused a spool format error. I have fixed the spool reading code. A
6339 related problem was that the "from" clause in the Received: line became
6340 illegal because of the space(s). It is now covered by ${quote_local_part.
6342 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6345 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6347 83. Patch from Sieve maintainer to fix unterminated string problem in
6348 "vacation" handling.
6350 84. Some minor changes to the Linux configuration files to help with other
6351 OS variants using glibc.
6353 85. One more patch for Sieve to update vacation handling to latest spec.
6356 ----------------------------------------------------
6357 See the note above about the 4.44 and 4.50 releases.
6358 ----------------------------------------------------
6364 1. Change 4.43/35 introduced a bug that caused file counts to be
6365 incorrectly computed when quota_filecount was set in an appendfile
6368 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6369 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6372 3. Give more explanation in the error message when the command for a transport
6373 filter fails to execute.
6375 4. There are several places where Exim runs a non-Exim command in a
6376 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6377 was being done only for the command run by the queryprogram router. It is
6378 now done for all such subprocesses. The other cases are: ${run, transport
6379 filters, and the commands run by the lmtp and pipe transports.
6381 5. Some older OS have a limit of 256 on the maximum number of file
6382 descriptors. Exim was using setrlimit() to set 1000 as a large value
6383 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6384 systems. I've change it so that if it can't get 1000, it tries for 256.
6386 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6387 was an oversight, and furthermore, ever since the addition of extra
6388 controls (e.g. 4.43/32), the checks on when to allow different forms of
6389 "control" were broken. There should now be diagnostics for all cases when a
6390 control that does not make sense is encountered.
6392 7. $recipients is now available in the predata ACL (oversight).
6394 8. Tidy the search cache before the fork to do a delivery from a message
6395 received from the command line. Otherwise the child will trigger a lookup
6396 failure and thereby defer the delivery if it tries to use (for example) a
6397 cached ldap connection that the parent has called unbind on.
6399 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6400 of $address_data from the recipient verification was clobbered by the
6401 sender verification.
6403 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6404 was its contents. (It was OK if the option was not defined at all.)
6406 11. A "Completed" log line is now written for messages that are removed from
6407 the spool by the -Mrm option.
6409 12. $host_address is now set to the target address during the checking of
6410 ignore_target_hosts.
6412 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6413 being passed; this would have caused $sender_host_name to have been used if
6414 matching the list had actually called for a host name (not very likely,
6415 since this list is usually IP addresses). A host name is now passed as
6418 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6419 code that specifies a non-restarting handler (typically sigaction() in
6420 modern systems) in an attempt to fix a rare and obscure crash bug.
6422 15. Narrowed the window for a race in the daemon that could cause it to ignore
6423 SIGCHLD signals. This is not a major problem, because they are used only to
6424 wake it up if nothing else does.
6426 16. A malformed maildirsize file could cause Exim to calculate negative values
6427 for the mailbox size or file count. Odd effects could occur as a result.
6428 The maildirsize information is now recalculated if the size or filecount
6431 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6432 support for a long time. Removed HAVE_SYS_VFS_H.
6434 18. Updated exipick to current release from John Jetmore.
6436 19. Allow an empty sender to be matched against a lookup in an address list.
6437 Previously the only cases considered were a regular expression, or an
6440 20. Exim went into a mad DNS lookup loop when doing a callout where the
6441 host was specified on the transport, if the DNS lookup yielded more than
6444 21. The RFC2047 encoding function was originally intended for short strings
6445 such as real names; it was not keeping to the 75-character limit for
6446 encoded words that the RFC imposes. It now respects the limit, and
6447 generates multiple encoded words if necessary. To be on the safe side, I
6448 have increased the buffer size for the ${rfc2047: expansion operator from
6451 22. Failure to deliver a bounce message always caused it to be frozen, even if
6452 there was an errors_to setting on the router. The errors_to setting is now
6455 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6456 canonical form (fully expanded) before being placed in
6457 $sender_host_address.
6459 24. Updated eximstats to version 1.33
6461 25. Include certificate and key file names in error message when GnuTLS fails
6462 to set them up, because the GnuTLS error message doesn't include the name
6463 of the failing file when there is a problem reading it.
6465 26. Expand error message when OpenSSL has problems setting up cert/key files.
6468 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6469 (this can affect the format of dates).
6471 28. exim_tidydb, when checking for the continued existence of a message for
6472 which it has found a message-specific retry record, was not finding
6473 messages that were in split spool directories. Consequently, it was
6474 deleting retry records that should have stayed in existence.
6476 29. eximstats updated to version 1.35
6477 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6478 1.35 - bugfix such that pie charts by volume are generated correctly
6480 30. The SPA authentication driver was not abandoning authentication and moving
6481 on to the next authenticator when an expansion was forced to fail,
6482 contradicting the general specification for all authenticators. Instead it
6483 was generating a temporary error. It now behaves as specified.
6485 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6486 (the order specifies the preference for clients). The order is now AES256,
6487 AES128, 3DES, ARCFOUR128.
6489 31. Small patch to Sieve code - explicitly set From: when generating an
6492 32. Exim crashed if a remote delivery caused a very long error message to be
6493 recorded - for instance if somebody sent an entire SpamAssassin report back
6494 as a large number of 550 error lines. This bug was coincidentally fixed by
6495 increasing the size of one of Exim's internal buffers (big_buffer) that
6496 happened as part of the Exiscan merge. However, to be on the safe side, I
6497 have made the code more robust (and fixed the comments that describe what
6500 33. Some experimental protocols are using DNS PTR records for new purposes. The
6501 keys for these records are domain names, not reversed IP addresses. The
6502 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6503 leaves it alone. Component reversal etc. now happens only for IP addresses.
6506 34. The host_aton() function is supposed to be passed a string that is known
6507 to be a valid IP address. However, in the case of IPv6 addresses, it was
6508 not checking this. This is a hostage to fortune. Exim now panics and dies
6509 if the condition is not met. A case was found where this could be provoked
6510 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6511 components; fortuitously, this particular loophole had already been fixed
6512 by change 4.50/55 or 4.44/33 above.
6514 If there are any other similar loopholes, the new check in host_aton()
6515 itself should stop them being exploited. The report I received stated that
6516 data on the command line could provoke the exploit when Exim was running as
6517 exim, but did not say which command line option was involved. All I could
6518 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6519 running as the user.
6522 35. There was a buffer overflow vulnerability in the SPA authentication code
6523 (which came originally from the Samba project). I have added a test to the
6524 spa_base64_to_bits() function which I hope fixes it.
6527 36. The daemon start-up calls getloadavg() while still root for those OS that
6528 need the first call to be done as root, but it missed one case: when
6529 deliver_queue_load_max is set with deliver_drop_privilege. This is
6530 necessary for the benefit of the queue runner, because there is no re-exec
6531 when deliver_drop_privilege is set.
6533 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6534 was in use, was not putting the data itself into the right store pool;
6535 consequently, it could be overwritten for a subsequent message in the same
6536 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6539 38. Sometimes the final signoff response after QUIT could fail to get
6540 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6541 < 0 before doing a fflush(). This bug looks as though it goes back to the
6542 introduction of TLS in release 3.20, but "sometimes" must have been rare
6543 because the tests only now provoked it.
6549 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6550 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6551 Now it returns an integer. A number of calls were still expecting a T/F
6552 return. Fortuitously, in all cases, the tests worked in OK situations,
6553 which is the norm. However, things would have gone wrong on any write
6554 failures on the smtp file descriptor. This function is used when sending
6555 messages over SMTP and also when doing verify callouts.
6557 2. When Exim is called to do synchronous delivery of a locally submitted
6558 message (the -odf or -odi options), it no longer closes stderr before doing
6561 3. Implemented the mua_wrapper option.
6563 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6565 5. Implemented the functions header_remove(), header_testname(),
6566 header_add_at_position(), and receive_remove_recipient(), and exported them
6569 6. If an ACL "warn" statement specified the addition of headers, Exim already
6570 inserted X-ACL-Warn: at the start if there was no header name. However, it
6571 was not making this test for the second and subsequent header lines if
6572 there were newlines in the string. This meant that an invalid header could
6573 be inserted if Exim was badly configured.
6575 7. Allow an ACL "warn" statement to add header lines at the start or after all
6576 the Received: headers, as well as at the end.
6578 8. Added the rcpt_4xx retry error code.
6580 9. Added postmaster_mailfrom=xxx to callout verification option.
6582 10. Added mailfrom=xxxx to the callout verification option, for verify=
6585 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6586 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6588 12. Inserted some casts to stop certain compilers warning when using pointer
6589 differences as field lengths or precisions in printf-type calls (mostly
6590 affecting debugging statements).
6592 13. Added optional readline() support for -be (dynamically loaded).
6594 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6595 same clock tick as a message's arrival, so that its received time was the
6596 same as the "first fail" time on the retry record, and that message
6597 remained on the queue past the ultimate address timeout, every queue runner
6598 would try a delivery (because it was past the ultimate address timeout) but
6599 after another failure, the ultimate address timeout, which should have then
6600 bounced the address, did not kick in. This was a "< instead of <=" error;
6601 in most cases the first failure would have been in the next clock tick
6602 after the received time, and all would be well.
6604 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6605 being recognized when the domain list was tested by the match_domain
6606 condition in an expansion string.
6608 16. Added the ${str2b64: operator.
6610 17. Exim was always calling setrlimit() to set a large limit for the number of
6611 processes, without checking whether the existing limit was already
6612 adequate. (It did check for the limit on file descriptors.) Furthermore,
6613 errors from getrlimit() and setrlimit() were being ignored. Now they are
6614 logged to the main and panic logs, but Exim does carry on, to try to do its
6615 job under whatever limits there are.
6617 18. Imported PCRE 5.0.
6619 19. Trivial typo in log message " temporarily refused connection" (the leading
6622 20. If the log selector return_path_on_delivery was set and an address was
6623 redirected to /dev/null, the delivery process crashed because it assumed
6624 that a return path would always be set for a "successful" delivery. In this
6625 case, the whole delivery is bypassed as an optimization, and therefore no
6628 21. Internal re-arrangement: the function for sending a challenge and reading
6629 a response while authentication was assuming a zero-terminated challenge
6630 string. It's now changed to take a pointer and a length, to allow for
6631 binary data in such strings.
6633 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6635 23. Exim was not respecting finduser_retries when seeking the login of the
6636 uid under which it was called; it was always trying 10 times. (The default
6637 setting of finduser_retries is zero.) Also, it was sleeping after the final
6638 failure, which is pointless.
6640 24. Implemented tls_on_connect_ports.
6642 25. Implemented acl_smtp_predata.
6644 26. If the domain in control=submission is set empty, Exim assumes that the
6645 authenticated id is a complete email address when it generates From: or
6646 Sender: header lines.
6648 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6649 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6650 chown and chgrp in /bin and hostname in /usr/bin.
6652 28. Exim was keeping the "process log" file open after each use, just as it
6653 does for the main log. This opens the possibility of it remaining open for
6654 long periods when the USR1 signal hits a daemon. Occasional processlog
6655 errors were reported, that could have been caused by this. Anyway, it seems
6656 much more sensible not to leave this file open at all, so that is what now
6659 29. The long-running daemon process does not normally write to the log once it
6660 has entered its main loop, and it closes the log before doing so. This is
6661 so that log files can straightforwardly be renamed and moved. However,
6662 there are a couple of unusual error situations where the daemon does write
6663 log entries, and I had neglected to close the log afterwards.
6665 30. The text of an SMTP error response that was received during a remote
6666 delivery was being truncated at 512 bytes. This is too short for some of
6667 the long messages that one sometimes sees. I've increased the limit to
6670 31. It is now possible to make retry rules that apply only when a message has a
6671 specific sender, in particular, an empty sender.
6673 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6674 it possible to be selective about when SMTP synchronization is enforced.
6676 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6678 32. Implemented hosts_connection_nolog.
6680 33. Added an ACL for QUIT.
6682 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6685 35. Added mailbox_size and mailbox_filecount to appendfile.
6687 36. Added control = no_multiline_responses to ACLs.
6689 37. There was a bug in the logic of the code that waits for the clock to tick
6690 in the case where the clock went backwards by a substantial amount such
6691 that the microsecond fraction of "now" was more than the microsecond
6692 fraction of "then" (but the whole seconds number was less).
6694 38. Added support for the libradius Radius client library this is found on
6695 FreeBSD (previously only the radiusclient library was supported).
6701 1. When certain lookups returned multiple values in the form name=value, the
6702 quoting of the values was not always being done properly. Specifically:
6703 (a) If the value started with a double quote, but contained no whitespace,
6705 (b) If the value contained whitespace other than a space character (i.e.
6706 tabs or newlines or carriage returns) it was not quoted.
6707 This fix has been applied to the mysql and pgsql lookups by writing a
6708 separate quoting function and calling it from the lookup code. The fix
6709 should probably also be applied to nisplus, ibase and oracle lookups, but
6710 since I cannot test any of those, I have not disturbed their existing code.
6712 2. A hit in the callout cache for a specific address caused a log line with no
6713 reason for rejecting RCPT. Now it says "Previous (cached) callout
6714 verification failure".
6716 3. There was an off-by-one bug in the queryprogram router. An over-long
6717 return line was truncated at 256 instead of 255 characters, thereby
6718 overflowing its buffer with the terminating zero. As well as fixing this, I
6719 have increased the buffer size to 1024 (and made a note to document this).
6721 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6722 when Exim is waiting for an SMTP response from a remote server, Exim
6723 restarts its select() call on the socket, thereby resetting its timeout.
6724 This is not a problem when such interrupts are rare. Somebody set up a cron
6725 job to run exiwhat every 2 minutes, which is less than the normal select()
6726 timeout (5 or 10 minutes). This meant that the select() timeout never
6727 kicked in because it was always reset. I have fixed this by comparing the
6728 time when an interrupt arrives with the time at the start of the first call
6729 to select(). If more time than the timeout has elapsed, the interrupt is
6730 treated as a timeout.
6732 5. Some internal re-factoring in preparation for the addition of Sieve
6733 extensions (by MH). In particular, the "personal" test is moved to a
6734 separate function, and given an option for scanning Cc: and Bcc: (which is
6735 not set for Exim filters).
6737 6. When Exim created an email address using the login of the caller as the
6738 local part (e.g. when creating a From: or Sender: header line), it was not
6739 quoting the local part when it contained special characters such as @.
6741 7. Installed new OpenBSD configuration files.
6743 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6744 try to make them clearer.
6746 9. Callout options, other than the timeout value, were being ignored when
6747 verifying sender addresses in header lines. For example, when using
6749 verify = header_sender/callout=no_cache
6751 the cache was (incorrectly) being used.
6753 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6754 only the Cygwin environment.
6756 11. When return_path_on_delivery was set as a log selector, if different remote
6757 addresses in the same message used different return paths and parallel
6758 remote delivery occurred, the wrong values would sometimes be logged.
6759 (Whenever a remote delivery process finished, the return path value from
6760 the most recently started remote delivery process was logged.)
6762 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6763 lines when AUTH and/or TLS are in use. This is the "received protocol"
6764 field. Exim used to use "asmtp" for authenticated SMTP, without any
6765 indication (in the protocol name) for TLS use. Now it follows the RFC and
6766 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6767 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6768 names appear in log lines as well as in Received: header lines.
6770 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6771 extensions, and comparison tests, and to fix some bugs.
6773 14. Changes to the "personal" filter test:
6775 (1) The test was buggy in that it was just doing the equivalent of
6776 "contains" tests on header lines. For example, if a user's address was
6777 anne@some.where, the "personal" test would incorrectly be true for
6779 To: susanne@some.where
6781 This test is now done by extracting each address from the header in turn,
6782 and checking the entire address. Other tests that are part of "personal"
6783 are now done using regular expressions (for example, to check local parts
6784 of addresses in From: header lines).
6786 (2) The list of non-personal local parts in From: addresses has been
6787 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6788 taken from the Sieve specification recommendations.
6790 (3) If the message contains any header line starting with "List-" it is
6791 treated as non-personal.
6793 (4) The test for "circular" in the Subject: header line has been removed
6794 because it now seems ill-conceived.
6796 15. Minor typos in src/EDITME comments corrected.
6798 16. Installed latest exipick from John Jetmore.
6800 17. If headers_add on a router specified a text string that was too long for
6801 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6802 of string_sprintf() is now avoided.
6804 18. $message_body_size was not set (it was always zero) when running the DATA
6805 ACL and the local_scan() function.
6807 19. For the "mail" command in an Exim filter, no default was being set for
6808 the once_repeat time, causing a random time value to be used if "once" was
6809 specified. (If the value happened to be <= 0, no repeat happened.) The
6810 default is now 0s, meaning "never repeat". The "vacation" command was OK
6811 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6812 (I found it when inspecting the code).
6814 20. There is now an overall timeout for performing a callout verification. It
6815 defaults to 4 times the callout timeout, which applies to individual SMTP
6816 commands during the callout. The overall timeout applies when there is more
6817 than one host that can be tried. The timeout is checked before trying the
6818 next host. This prevents very long delays if there are a large number of
6819 hosts and all are timing out (e.g. when the network connections are timing
6820 out). The value of the overall timeout can be changed by specifying an
6821 additional sub-option for "callout", called "maxwait". For example:
6823 verify = sender/callout=5s,maxwait=20s
6825 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6826 to the end before writing, but this should make it even safer).
6828 22. Exim was forgetting that it had advertised PIPELINING for the second and
6829 subsequent messages on an SMTP connection. It was also not resetting its
6830 memory on STARTTLS and an internal HELO.
6832 23. When Exim logs an SMTP synchronization error within a session, it now
6833 records whether PIPELINING has been advertised or not.
6835 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6836 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6837 rather than long int.
6839 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6841 26. Added the never_mail option to autoreply.
6847 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6848 crash if the getsockname() call failed; this can happen if a connection is
6849 closed very soon after it is established. The problem was simply in the
6850 order in which certain operations were done, causing Exim to try to write
6851 to the SMTP stream before it had set up the file descriptor. The bug has
6852 been fixed by making things happen in the correct order.
6858 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6859 before the connection was closed, thus losing the rejection response.
6861 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6862 some early Solaris releases, but causes trouble in current releases where
6863 socklen_t is defined.
6865 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6868 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6871 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6872 that releasing the top of it at the end releases what was used for sub-
6873 expansions (unless the block got too big). However, discard this block if
6874 the first thing is a variable or header, so that we can use its block when
6875 it is dynamic (useful for very large $message_headers, for example).
6877 6. Lookups now cache *every* query, not just the most recent. A new, separate
6878 store pool is used for this. It can be recovered when all lookup caches are
6879 flushed. Lookups now release memory at the end of their result strings.
6880 This has involved some general refactoring of the lookup sources.
6882 7. Some code has been added to the store_xxx() functions to reduce the amount
6883 of flapping under certain conditions.
6885 8. log_incoming_interface used to affect only the <= reception log lines. Now
6886 it causes the local interface and port to be added to several more SMTP log
6887 lines, for example "SMTP connection from", and rejection lines.
6889 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6891 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6893 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6894 could be overwritten at the end of the current message (or the start of a
6895 new message if it was set in a HELO ACL). The value is now preserved for
6896 the duration of the SMTP connection.
6898 12. If a transport had a headers_rewrite setting, and a matching header line
6899 contained an unqualified address, that address was qualified, even if it
6900 did not match any rewriting rules. The underlying bug was that the values
6901 of the flags that permit the existence of unqualified sender and recipient
6902 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6903 non-local messages, and by -bnq for local messages) were not being
6904 preserved with the message after it was received.
6906 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6907 "next input=" as part of the text comprising the host identity instead of
6908 the correct text. The code was using the same buffer for two different
6909 strings. However, depending on which order the printing function evaluated
6910 its arguments, the bug did not always show up. Under Linux, for example, my
6911 test suite worked just fine.
6913 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6914 doesn't work with some older versions of Perl. It has been changed to "my",
6915 which in any case is probably the better facility to use.
6917 15. A really picky compiler found some instances of statements for creating
6918 error messages that either had too many or two few arguments for the format
6921 16. The size of the buffer for calls to the DNS resolver has been increased
6922 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6923 for addresses that have a lot of PTR records. This alleviates a problem; it
6924 does not fully solve it.
6926 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6927 buffer now truncates the list and logs the incident, which is the same
6928 action as happens when Exim is looking up a host name and its aliases.
6929 Previously in this situation something unpredictable would happen;
6930 sometimes it was "internal error: store_reset failed".
6932 18. If a server dropped the connection unexpectedly when an Exim client was
6933 using GnuTLS and trying to read a response, the client delivery process
6934 crashed while trying to generate an error log message.
6936 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6937 string, for example:
6939 warn message = H1: something\nH2: something
6941 the text was added as a single header line from Exim's point of view
6942 though it ended up OK in the delivered message. However, searching for the
6943 second and subsequent header lines using $h_h2: did not work. This has been
6944 fixed. Similarly, if a system filter added multiple headers in this way,
6945 the routers could not see them.
6947 20. Expanded the error message when iplsearch is called with an invalid key to
6948 suggest using net-iplsearch in a host list.
6950 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6951 ACLs are no longer actually imposed (and a message to that effect is
6954 22. If a "gecos" field in a passwd entry contained escaped characters, in
6955 particular, if it contained a \" sequence, Exim got it wrong when building
6956 a From: or a Sender: header from that name. A second bug also caused
6957 incorrect handling when an unquoted " was present following a character
6958 that needed quoting.
6960 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6961 was not being matched caselessly.
6963 24. Arranged for all hyphens in the exim.8 source to be escaped with
6966 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6967 properly. Recipient callout cache records were still being keyed to include
6968 the sender, even when use_sender was set false. This led to far more
6969 callouts that were necessary. The sender is no longer included in the key
6970 when use_sender is false.
6972 26. Added "control = submission" modifier to ACLs.
6974 27. Added the ${base62d: operator to decode base 62 numbers.
6976 28. dnsdb lookups can now access SRV records.
6978 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6979 the configuration file.
6981 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6982 (-v) mode. This makes the output for a verbose queue run more intelligible.
6984 31. Added a use_postmaster feature to recipient callouts.
6986 32. Added the $body_zerocount variable, containing the number of binary zero
6987 bytes in the message body.
6989 33. The time of last modification of the "new" subdirectory is now used as the
6990 "mailbox time last read" when there is a quota error for a maildir
6993 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6995 35. Added +ignore_unknown as a special item in host lists.
6997 36. Code for decoding IPv6 addresses in host lists is now included, even if
6998 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6999 address was recognized as an IP address, but was then not correctly decoded
7000 into binary, causing unexpected and incorrect effects when compared with
7007 1. Very minor rewording of debugging text in manualroute to say "list of
7008 hosts" instead of "hostlist".
7010 2. If verify=header_syntax was set, and a header line with an unqualified
7011 address (no domain) and a large number of spaces between the end of the
7012 name and the colon was received, the reception process suffered a buffer
7013 overflow, and (when I tested it) crashed. This was caused by some obsolete
7014 code that should have been removed. The fix is to remove it!
7016 3. When running in the test harness, delay a bit after writing a bounce
7017 message to get a bit more predictability in the log output.
7019 4. Added a call to search_tidyup() just before forking a reception process. In
7020 theory, someone could use a lookup in the expansion of smtp_accept_max_
7021 per_host which, without the tidyup, could leave open a database connection.
7023 5. Added the variables $recipient_data and $sender_data which get set from a
7024 lookup success in an ACL "recipients" or "senders" condition, or a router
7025 "senders" option, similar to $domain_data and $local_part_data.
7027 6. Moved the writing of debug_print from before to after the "senders" test
7030 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7031 problems for message scanning, either using a data ACL, or using
7032 local_scan() because the Received: header was not generated till after they
7033 were called (in order to set the time as the time of reception completion).
7034 I have revised the way this works. The header is now generated after the
7035 body is received, but before the ACL or local_scan() are called. After they
7036 are run, the timestamp in the header is updated.
7042 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7043 before starting a queue runner without re-exec. This happened only when
7044 deliver_drop_privilege was set or when the Exim user was set to root. The
7045 effect of the bug was that timeouts during subsequent deliveries caused
7046 crashes instead of being properly handled. The handler is now left at its
7047 default (and expected) setting.
7049 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7050 message, again when deliver_drop_privilege is set or Exim is run as root.
7051 The bug described in (1) was not present in this case, but the tidying up
7052 of the other signals was missing. I have made the two cases consistent.
7054 3. The ignore_target_hosts setting on a manualroute router was being ignored
7055 for hosts that were looked up using the /MX notation.
7057 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7060 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7061 operated on the sender address. After changing the $sender_address to <>
7062 for the sender address verify, Exim was re-instated it as the original
7063 (before rewriting) address, but remembering that it had rewritten it, so it
7064 wasn't rewriting it again. This bug also had the effect of breaking the
7065 sender address verification caching when the sender address was rewritten.
7067 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7068 This has been changed so that if the ip literal address matches
7069 ignore_target_hosts, the router declines.
7071 7. Added expansion conditions match_domain, match_address, and match_local_
7072 part (NOT match_host).
7074 8. The placeholder for the Received: header didn't have a length field set.
7076 9. Added code to Exim itself and to exim_lock to test for a specific race
7077 condition that could lead to file corruption when using MBX delivery. The
7078 issue is with the lockfile that is created in /tmp. If this file is removed
7079 after a process has opened it but before that process has acquired a lock,
7080 there is the potential for a second process to recreate the file and also
7081 acquire a lock. This could lead to two Exim processes writing to the file
7082 at the same time. The added code performs the same test as UW imapd; it
7083 checks after acquiring the lock that its file descriptor still refers to
7084 the same named file.
7086 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7087 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7090 11. Added the smtp_active_hostname option. If used, this will typically be made
7091 to depend on the incoming interface address. Because $interface_address is
7092 not set up until the daemon has forked a reception process, error responses
7093 that can happen earlier (such as "too many connections") no longer contain
7096 12. If an expansion in a condition on a "warn" statement fails because a lookup
7097 defers, the "warn" statement is abandoned, and the next ACL statement is
7098 processed. Previously this caused the whole ACL to be aborted.
7100 13. Added the iplsearch lookup type.
7102 14. Added ident_timeout as a log selector.
7104 15. Added tls_certificate_verified as a log selector.
7106 16. Added a global option tls_require_ciphers (compare the smtp transport
7107 option of the same name). This controls incoming TLS connections.
7109 17. I finally figured out how to make tls_require_ciphers do a similar thing
7110 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7111 before starting the TLS session.
7113 18. Tabs are now shown as \t in -bP output.
7115 19. If the log selector return_path_on_delivery was set, Exim crashed when
7116 bouncing a message because it had too many Received: header lines.
7118 20. If two routers both had headers_remove settings, and the first one included
7119 a superfluous trailing colon, the final name in the first list and the
7120 first name in the second list were incorrectly joined into one item (with a
7121 colon in the middle).
7127 1. Added -C and -D options to the exinext utility, mainly to make it easier
7128 to include in the automated testing, but these could be helpful when
7129 multiple configurations are in use.
7131 2. The exinext utility was not formatting the output nicely when there was
7132 an alternate port involved in the retry record key, nor when there was a
7133 message id as well (for retries that were specific to a specific message
7134 and a specific host). It was also confused by IPv6 addresses, because of
7135 the additional colons they contain. I have fixed the IPv4 problem, and
7136 patched it up to do a reasonable job for IPv6.
7138 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7139 delivery, the log line now contains "pipelined" if PIPELINING was used.
7141 4. An SMTP transport process used to panic and die if the bind() call to set
7142 an explicit outgoing interface failed. This has been changed; it is now
7143 treated in the same way as a connect() failure.
7145 5. A reference to $sender_host_name in the part of a conditional expansion
7146 that was being skipped was still causing a DNS lookup. This no longer
7149 6. The def: expansion condition was not recognizing references to header lines
7150 that used bh_ and bheader_.
7152 7. Added the _cache feature to named lists.
7154 8. The code for checking quota_filecount in the appendfile transport was
7155 allowing one more file than it should have been.
7157 9. For compatibility with Sendmail, the command line option
7165 and sets the incoming protocol and host name (for trusted callers). The
7166 host name and its colon can be omitted when only the protocol is to be set.
7167 Note the Exim already has two private options, -pd and -ps, that refer to
7168 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7169 "s", but I don't think that's a major issue.
7171 10. A number of refactoring changes to the code, none of which should affect
7174 (a) The number of logging options was getting close to filling up the
7175 32-bit word that was used as a bit map. I have split them into two classes:
7176 those that are passed in the argument to log_write(), and those that are
7177 only ever tested independently outside of that function. These are now in
7178 separate 32-bit words, so there is plenty of room for expansion again.
7179 There is no change in the user interface or the logging behaviour.
7181 (b) When building, for example, log lines, the code previously used a
7182 macro that called string_cat() twice, in order to add two strings. This is
7183 not really sufficiently general. Furthermore, there was one instance where
7184 it was actually wrong because one of the argument was used twice, and in
7185 one call a function was used. (As it happened, calling the function twice
7186 did not affect the overall behaviour.) The macro has been replaced by a
7187 function that can join an arbitrary number of extra strings onto a growing
7190 (c) The code for expansion conditions now uses a table and a binary chop
7191 instead of a serial search (which was left over from when there were very
7192 few conditions). Also, it now recognizes conditions like "pam" even when
7193 the relevant support is not compiled in: a suitably worded error message is
7194 given if an attempt is made to use such a condition.
7196 11. Added ${time_interval:xxxxx}.
7198 12. A bug was causing one of the ddress fields not to be passed back correctly
7199 from remote delivery subprocesses. The field in question was not being
7200 subsequently used, so this caused to problems in practice.
7202 13. Added new log selectors queue_time and deliver_time.
7204 14. Might have fixed a bug in maildirsizefile handling that threw up
7205 "unexpected character" debug warnings, and recalculated the data
7206 unnecessarily. In any case, I expanded the warning message to give more
7209 15. Added the message "Restricted characters in address" to the statements in
7210 the default ACL that block characters like @ and % in local parts.
7212 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7213 Three changes have been made:
7215 (a) There was a serious bug; a negative response to MAIL caused the whole
7216 recipient domain to be cached as invalid, thereby blocking all messages
7217 to all local parts at the same domain, from all senders. This bug has
7218 been fixed. The domain is no longer cached after a negative response to
7219 MAIL if the sender used is not empty.
7221 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7224 (c) A new callout option, "use_sender" has been added for people who want
7225 the modified behaviour.
7231 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7234 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7235 indeed breaks things for older releases.
7237 3. Added additional logging to the case where there is a problem reading data
7238 from a filter that is running in a subprocess using a pipe, in order to
7239 try to track down a specific problem.
7241 4. Testing facility fudge: when running in the test harness and attempting
7242 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7243 getting "No route to host". Convert this to a timeout.
7245 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7248 6. Some OS don't have socklen_t but use size_t instead. This affects the
7249 fifth argument of getsockopt() amongst other things. This is now
7250 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7251 can be set for individual OS. I have set it for SunOS5, OSF1, and
7252 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7253 some earlier ones do not.
7255 7. Change 4.30/15 was not doing the test caselessly.
7257 8. The standard form for an IPv6 address literal was being rejected by address
7258 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7259 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7260 this, as well as the form without the "IPv6" on the front (but only when
7261 address literals are enabled, of course).
7263 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7265 10. Exim crashed if a message with an empty sender address specified by -f
7266 encountered a router with an errors_to setting. This could be provoked only
7267 by a command such as
7271 where an empty string was supplied; "<>" did not hit this bug.
7273 11. Installed PCRE release 4.5.
7275 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7276 remained set. It is now erased.
7278 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7279 times from message ids (which are base 36 rather than the normal 62).
7281 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7282 were being counted as actual protocol errors, and logged if the log
7283 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7284 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7285 and DATA following a set of rejected RCPTs do not count as protocol errors.
7286 In other words, Exim assumes they were pipelined, though this may not
7287 actually be the case. Of course, in all cases the client gets an
7288 appropriate error code.
7290 15. If a lookup fails in an ACL condition, a message about the failure may
7291 be available; it is used if testing the ACL cannot continue, because most
7292 such messages specify what the cause of the deferral is. However, some
7293 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7294 that caused an old message to be retained and used if a later statement
7295 caused a defer, replacing the real cause of the deferral.
7297 16. If an IP address had so many PTR records that the DNS lookup buffer
7298 was not large enough to hold them, Exim could crash while trying to process
7299 the truncated data. It now detects and logs this case.
7301 17. Further to 4.21/58, another change has been made: if (and only if) the
7302 first line of a message (the first header line) ends with CRLF, a bare LF
7303 in a subsequent header line has a space inserted after it, so as not to
7304 terminate the header.
7306 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7307 unnecessarily, used atoi() instead of strtol(), and didn't check the
7308 termination when getting file sizes from file names by regex.
7310 19. Completely re-implemented the support for maildirsize files, in the light
7311 of a number of problems with the previous contributed implementation
7312 (4.30/29). In particular:
7314 . If the quota is zero, the maildirsize file is maintained, but no quota is
7317 . If the maildir directory does not exist, it is created before any attempt
7318 to write a maildirsize file.
7320 . The quota value in the file is just a cache; if the quota is changed in
7321 the transport, the new value overrides.
7323 . A regular expression is available for excluding directories from the
7326 20. The autoreply transport checks the characters in options that define the
7327 message's headers; it allows continued headers, but it was checking with
7328 isspace() after an embedded newline instead of explicitly looking for a
7331 21. If all the "regular" hosts to which an address was routed had passed their
7332 expiry times, and had not reached their retry times, the address was
7333 bounced, even if fallback hosts were defined. Now Exim should go on to try
7336 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7337 equivalent code in the SMTP transport. Some hosts send humungous responses
7338 to HELO/EHLO, more than 1024 it seems.
7340 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7341 gives compiler warnings in some environments. I've now done it "properly",
7344 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7345 (because of problems with the built-in one) was declared to return uschar *
7346 instead of char *, causing compiler failure.
7348 25. Fixed a file descriptor leak when processing alias/forward files.
7350 26. Fixed a minor format string issue in dbfn.c.
7352 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7354 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7355 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7358 29. When a sender address is verified, it is cached, to save repeating the test
7359 when there is more than one recipient in a message. However, when the
7360 verification involves a callout, it is possible for different callout
7361 options to be set for different recipients. It is too complicated to keep
7362 track of this in the cache, so now Exim always runs a verification when a
7363 callout is required, relying on the callout cache for the optimization.
7364 The overhead is duplication of the address routing, but this should not be
7367 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7368 to be verified with callout=postmaster, and the main callout worked but the
7369 postmaster check failed, the verification correctly failed. However, if a
7370 subsequent RCPT command asked for sender verification *without* the
7371 postmaster check, incorrect caching caused this verification also to fail,
7374 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7375 it was not caching the DNS options (qualify_single, search_parents) that
7376 were used when the lookup failed. A subsequent lookup with different
7377 options therefore always gave the same answer, though there were cases
7378 where it should not have. (Example: a "domains = !$mx_any" option on a
7379 dnslookup router: the "domains" option is always processed without any
7380 widening, but the router might have qualify_single set.) Now Exim uses the
7381 cached value only when the same options are set.
7383 32. Added John Jetmore's "exipick" utility to the distribution.
7385 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7386 than a timeout (e.g. a certificate is required, and is not provided), an
7387 Exim server now closes the connection immediately. Previously it waited for
7388 the client to close - but if the client is SSL, it seems that they each
7389 wait for each other, leading to a delay before one of them times out.
7391 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7392 maintained 0.8.x compatibility because I don't think many are using it, and
7393 it is clearly obsolete.
7395 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7398 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7399 client certificate was expired. A simple patch fixes this, though I don't
7400 understand the full logic of why the verify callback is called multiple
7403 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7404 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7405 which causes problems with some clients (such as the Certicom SSL Plus
7406 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7407 disables the coutermeasure allowing Eudora to connect."
7409 38. Exim was not checking that a write() to a log file succeeded. This could
7410 lead to Bad Things if a log got too big, in particular if it hit a file
7411 size limit. Exim now panics and dies if it cannot write to a log file, just
7412 as it does if it cannot open a log file.
7414 39. Modified OS/Makefile-Linux so that it now contains
7416 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7418 The two -D definitions ensure that Exim is compiled with large file
7419 support, which makes it possible to handle log files that are bigger than
7422 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7423 instance) a domain was checked against a named list that involved a lookup,
7424 causing $domain_data to be set, then another domain was checked against the
7425 same list, then the first domain was re-checked, the value of $domain_data
7426 after the final check could be wrong. In particular, if the second check
7427 failed, it could be set empty. This bug probably also applied to
7430 41. The strip_trailing_dot option was not being applied to the address given
7431 with the -f command-line option.
7433 42. The code for reading a message's header from the spool was incrementing
7434 $received_count, but never initializing it. This meant that the value was
7435 incorrect (doubled) while delivering a message in the same process in which
7436 it was received. In the most common configuration of Exim, this never
7437 happens - a fresh exec is done - but it can happen when
7438 deliver_drop_privilege is set.
7440 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7441 it now includes up to 150 characters of the unexpected data in the log
7444 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7445 and building data strings. The size of both of these buffers was 10 000
7446 bytes - far larger than anybody would *ever* want, thought I. Needless to
7447 say, somebody hit the limit. I have increased the maximum line length to
7448 20 000 and the maximum data length of concatenated lines to 100 000. I have
7449 also fixed two bugs, because there was no checking on these buffers. Tsk,
7450 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7451 buffer is too small.
7453 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7454 lsearch lookups. Now it does.
7456 46. When parsing a route_list item in a manualroute router, a fixed-length
7457 buffer was used for the list of hosts. I made this 1024 bytes long,
7458 thinking that nobody would ever have a list of hosts that long. Wrong.
7459 Somebody had a whole pile of complicated expansion conditions, and the
7460 string was silently truncated, leading to an expansion error. It turns out
7461 that it is easier to change to an unlimited length (owing to other changes
7462 that have happened since this code was originally written) than to build
7463 structure for giving a limitation error. The length of the item that
7464 expands into the list of hosts is now unlimited.
7466 47. The lsearch lookup could not handle data where the length of text line was
7467 more than 4095 characters. Such lines were truncated, leading to shortened
7468 data being returned. It should now handle lines of any length.
7470 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7471 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7474 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7475 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7476 no longer happens in comment lines. A list of replacements is now placed
7477 at the head of all of the source files, except those whose only change is
7478 to replace PERL_COMMAND in the very first #! line.
7480 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7481 messages on the queue, with a bottom-up merge sort, using code contributed
7482 by Michael Haardt. This should make operations like -bp somewhat faster on
7483 large queues. It won't affect queue runners, except when queue_run_in_order
7486 51. Installed eximstats 1.31 in the distribution.
7488 52. Added support for SRV lookups to the dnslookup router.
7490 53. If an ACL referred to $message_body or $message_body_end, the value was not
7491 reset for any messages that followed in the same SMTP session.
7493 54. The store-handling optimization for building very long strings was not
7494 differentiating between the different store pools. I don't think this
7495 actually made any difference in practice, but I've tidied it.
7497 55. While running the routers to verify a sender address, $sender_address
7498 was still set to the sender address. This is wrong, because when routing to
7499 send a bounce to the sender, it would be empty. Therefore, I have changed
7500 it so that, while verifying a sender address, $sender_address is set to <>.
7501 (There is no change to what happens when verifying a recipient address.)
7503 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7504 target A or AAAA records (if not already returned) without resetting the
7505 qualify_single or search_parents options of the DNS resolver. These are
7506 inappropriate in this case because the targets of MX and SRV records must
7507 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7508 target that, when qualified, matched something in the local domain. These
7509 two options are now turned off when doing these lookups.
7511 57. It seems that at least some releases of Reiserfs (which does not have the
7512 concept of a fixed number of inodes) returns zero and not -1 for the
7513 number of available inodes. This interacted badly with check_spool_inodes,
7514 which assumed that -1 was the "no such thing" setting. What I have done is
7515 to check that the total number of inodes is greater than zero before doing
7516 the test of how many are available.
7518 58. When a "warn" ACL statement has a log_message modifier, the message is
7519 remembered, and not repeated. This is to avoid a lot of repetition when a
7520 message has many recipients that cause the same warning to be written.
7521 However, Exim was preserving the list of already written lines for an
7522 entire SMTP session, which doesn't seem right. The memory is now reset if a
7523 new message is started.
7525 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7526 debugging output unless log_rewrite was also set.
7528 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7529 of (int)(handle) when we know that handle contains (void *)(-1).
7531 61. The Exim daemon panic-logs an error return when it closes the incoming
7532 connection. However "connection reset by peer" seems to be common, and
7533 isn't really an error worthy of noting specially, so that particular error
7536 62. When Exim is trying to find all the local interfaces, it used to panic and
7537 die if the ioctl to get the interface flags failed. However, it seems that
7538 on at least one OS (Solaris 9) it is possible to have an interface that is
7539 included in the list of interfaces, but for which you get a failure error
7540 for this call. This happens when the interface is not "plumbed" into a
7541 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7542 failure of the "get flags" call assumes that the interface is down.
7544 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7545 makes life easier for people who are doing arithmetic on fields extracted
7546 from dates, where you often get leading zeros that should not be
7547 interpreted as octal.
7549 64. Added qualify_domain to the redirect router, to override the global
7552 65. If a pathologically long header line contained very many addresses (the
7553 report of this problem mentioned 10 000) and each of them was rewritten,
7554 Exim could use up a very large amount of memory. (It kept on making new
7555 copies of the header line as it rewrote, and never released the old ones.)
7556 At the expense of a bit more processing, the header rewriting function has
7557 been changed so that it no longer eats memory in this way.
7559 66. The generation of the Received: header has been moved from the time that a
7560 message starts to be received, to the time that it finishes. The timestamp
7561 in the Received: header should now be very close to that of the <= log
7562 line. There are two side-effects of this change:
7564 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7565 logged header lines no longer include the local Received: line, because
7566 it has not yet been created. The same applies to a copy of the message
7567 that is returned to a non-SMTP sender when a message is rejected.
7569 (b) When a filter file is tested using -bf, no additional Received: header
7570 is added to the test message. After some thought, I decided that this
7573 This change does not affect the value of $received_for. It is still set
7574 after address rewriting, but before local_scan() is called.
7576 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7578 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7579 gave an unhelpful panic error message, and a defer error. I have managed to
7580 change this behaviour so that it now rejects any supplied certificate,
7581 which seems right, as the list of acceptable certificates is empty.
7583 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7584 gave an unhelpful defer error. I have not managed to make this reject any
7585 supplied certificates, but the error message it gives is "no certificate
7586 supplied", which is not helpful.
7588 70. exigrep's output now also includes lines that are not associated with any
7589 message, but which match the given pattern. Implemented by a patch from
7590 Martin Sluka, which also tidied up the Perl a bit.
7592 71. Recipient callout verification, like sender verification, was using <> in
7593 the MAIL FROM command. This isn't really the right thing, since the actual
7594 sender may affect whether the remote host accepts the recipient or not. I
7595 have changed it to use the actual sender in the callout; this means that
7596 the cache record is now keyed on a recipient/sender pair, not just the
7597 recipient address. There doesn't seem to be a real danger of callout loops,
7598 since a callout by the remote host to check the sender would use <>.
7599 [SEE ABOVE: changed after hitting problems.]
7601 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7602 temporary errors. However, in the case of such a code being given after
7603 the end of a data transmission (i.e. after ".") Exim was failing to write
7604 a retry record for the message. (Yes, there was some broken host that was
7605 actually sending 8xx at this point.)
7607 73. An unknown lookup type in a host list could cause Exim to panic-die when
7608 the list was checked. (An example that provoked this was putting <; in the
7609 middle of a list instead of at the start.) If this happened during a DATA
7610 ACL check, a -D file could be left lying around. This kind of configuration
7611 error no longer causes Exim to die; instead it causes a defer error. The
7612 incident is still logged to the main and panic logs.
7614 74. Buglet left over from Exim 3 conversion. The message "too many messages
7615 in one connection" was written to the rejectlog but not the mainlog, except
7616 when address rewriting (yes!) was being logged.
7618 75. Added write_rejectlog option.
7620 76. When a system filter was run not as root (that is, when system_filter_user
7621 was set), the values of the $n variables were not being returned to the
7622 main process; thus, they were not subsequently available in the $sn
7625 77. Added +return_path_on_delivery log selector.
7627 78. A connection timeout was being treated differently from recipients deferred
7628 when testing hosts_max_try with a message that was older than the host's
7629 retry timeout. (The host should not be counted, thus allowing all hosts to
7630 be tried at least once before bouncing.) This may have been the cause of an
7631 occasionally reported bug whereby a message would remain on the queue
7632 longer than the retry timeout, but would be bounced if a delivery was
7633 forced. I say "may" because I never totally pinned down the problem;
7634 setting up timeout/retry tests is difficult. See also the next item.
7636 79. The ultimate address timeout was not being applied to errors that involved
7637 a combination of host plus message (for example, a timeout on a MAIL
7638 command). When an address resolved to a number of possible hosts, and they
7639 were not all tried for each delivery (e.g. because of hosts_max_try), a
7640 message could remain on the queue longer than the retry timeout.
7642 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7645 81. Fixed an obscure SMTP outgoing bug which required at least the following
7646 conditions: (a) there was another message waiting for the same server;
7647 (b) the server returned 5xx to all RCPT commands in the first message so
7648 that the message was not completed; (c) the server dropped the connection
7649 or gave a negative response to the RSET that Exim sends to abort the
7650 transaction. The observed case was a dropped connection after DATA that had
7651 been sent in pipelining mode. That is, the server had advertised PIPELINING
7652 but was not implementing it correctly. The effect of the bug was incorrect
7653 behaviour, such as trying another host, and this could lead to a crash.
7659 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7660 and daemon.c were passed as pointers to ints; they should have been
7661 pointers to socklen_t variables (which are typically unsigned ints).
7663 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7666 3. Fixed a really odd bug that affected only the testing scheme; patching a
7667 certain fixed string in the binary changed the value of another string that
7668 happened to be identical to the end of the original first string.
7670 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7671 name", it returns that address as the IP address. On some operating
7672 systems (e.g. Solaris), it also passes back the IP address string as the
7673 "host name". However, on others (e.g. Linux), it passes back an empty
7674 string. Exim wasn't checking for this, and was changing the host name to an
7675 empty string, assuming it had been canonicalized.
7677 5. Although rare, it is permitted to have more than one PTR record for a given
7678 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7679 all the names associated with an address, because they do in Solaris.
7680 However, it seems that they do not in Linux for data that comes from the
7681 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7682 I found this out when I moved to a new Linux workstation and tried to run
7683 the Exim test suite.
7685 To get round this problem I have changed the code so that it now does its
7686 own call to the DNS to look up PTR records when searching for a host name.
7687 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7688 addresses that are only in /etc/hosts are still found.
7690 This behaviour is, however, controlled by an option called host_lookup_
7691 order, which defaults to "bydns:byaddr". If people want to use the other
7692 order, or indeed, just use one or the other means of lookup, they can
7693 specify it in this variable.
7695 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7696 some operating systems, this comes back from gethostbyaddr() as an empty
7697 string, and this is what Exim used to test for. However, it seems that in
7698 other systems, "." is yielded. Exim now tests for this case too.
7700 7. The values of check_spool_space and check_log_space are now held internally
7701 as a number of kilobytes instead of an absolute number of bytes. If a
7702 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7703 kilobyte. This means that much larger values can be stored.
7705 8. Exim monitor: an attempt to get the action menu when not actually pointing
7706 at a message produces an empty menu entitled "No message selected". This
7707 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7708 no entries in it ("Shell widget menu has zero width and/or height"). So I
7709 have added a single, blank menu entry in this case.
7711 9. Added ${quote_local_part.
7713 10. MIME decoding is now applied to the contents of Subject: header lines when
7716 11. Now that a reference to $sender_host_address automatically causes a reverse
7717 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7718 host lookup before query-style lookups in lists that might use this
7719 variable. This has therefore been abolished, and the "net-" prefix is no
7720 longer necessary for query-style lookups.
7722 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7723 have been a typo for LFLAGS, so it has been changed.
7725 13. The install script calls Exim with "-C /dev/null" in order to find the
7726 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7727 to be output. However, since Exim outputs its version number before the
7728 error, it didn't break the script. It just looked ugly. I fixed this by
7729 always allowing "-C /dev/null" if the caller is root.
7731 14. Ignore overlarge ACL variable number when reading spool file - insurance
7732 against a later release with more variables having written the file.
7734 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7735 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7736 this, as well as the form without the "IPv6" on the front.
7738 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7739 OS/Makefile-Darwin file.
7741 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7742 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7744 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7745 variable. However, some OS use ints for inodes. I've added cast to long int
7746 to get rid of the compiler warning.
7748 19. I had forgotten to lock out "/../" in configuration file names when
7749 ALT_CONFIG_PREFIX was set.
7751 20. Routers used for verification do not need to specify transports. However,
7752 if such a router generated a host list, and callout was configured, Exim
7753 crashed, because it could not find a port number from the (non-existent)
7754 transport. It now assumes port 25 in this circumstance.
7756 21. Added the -t option to exigrep.
7758 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7759 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7760 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7761 them is set, LOOKUP_LSEARCH is forced.
7763 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7765 24. Added sender and host information to the "rejected by local_scan()" log
7766 line; previously there was no indication of these.
7768 25. Added .include_if_exists.
7770 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7771 Linux. It turns out that not all file systems support this. Apparently some
7772 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7773 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7774 not supported on the file descriptor, is now ignored when Exim is trying to
7775 sync a directory. This applies only to Linux.
7777 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7779 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7780 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7783 29. Added support for maildirsize files from supplied patch (modified a bit).
7785 30. The use of :fail: followed by an empty string could lead Exim to respond to
7786 sender verification failures with (e.g.):
7788 550 Verification failed for <xxx>
7789 550 Sender verify failed
7791 where the first response line was missing the '-' that indicates it is not
7792 the final line of the response.
7794 31. The loop for finding the name of the user that called Exim had a hardwired
7795 limit of 10; it now uses the value of finduser_retries, which is used for
7796 all other user lookups.
7798 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7801 33. Exim was neglecting to zero errno before one call of strtol() when
7802 expanding a string and expecting an integer value. On some systems this
7803 resulted in spurious "integer overflow" errors. Also, it was casting the
7804 result into an int without checking.
7806 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7807 did not work. The code looks as if it has *never* worked, though it appears
7808 to have been documented since at least release 1.62. I have made it work.
7810 35. The "timeout_DNS" error in retry rules, also documented since at least
7811 1.62, also never worked. As it isn't clear exactly what this means, and
7812 clearly it isn't a major issue, I have abolished the feature by treating it
7813 as "timeout", and writing a warning to the main and panic logs.
7815 36. The display of retry rules for -brt wasn't always showing the error code
7818 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7819 timeout_connect_A, timeout_connect_MX.
7821 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7822 to the empty sender.
7824 39. The daemon was not analysing the content of -oX till after it had closed
7825 stderr and disconnected from the controlling terminal. This meant that any
7826 syntax errors were only noted on the panic log, and the return code from
7827 the command was 0. By re-arranging the code a little, I've made the
7828 decoding happen first, so such errors now appear on stderr, and the return
7829 code is 1. However, the actual setting up of the sockets still happens in
7830 the disconnected process, so errors there are still only recorded on the
7833 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7834 connections (as happens on some IP stacks) was logged at start up time as
7835 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7836 it from "IPv6 and IPv4", which means that two separate sockets are being
7839 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7840 says whether AF_INET or AF_INET6 was passed as an argument.
7842 42. Exiwhat output was messed up when time zones were included in log
7845 43. Exiwhat now gives more information about the daemon's listening ports,
7846 and whether -tls-on-connect was used.
7848 44. The "port" option of the smtp transport is now expanded.
7850 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7851 silently ignored. Now an error message is written to the main and panic
7854 46. There's a new ACL modifier called "logwrite" which writes to a log file
7855 as soon as it is encountered.
7857 47. Added $local_user_uid and $local_user_gid at routing time.
7859 48. Exim crashed when trying to verify a sender address that was being
7862 49. Exim was recognizing only a space character after ".include". It now also
7863 recognizes a tab character.
7865 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7866 extracting the relevant information from the specification. The man page no
7867 longer contains scrambled data for the -d option, and I've added a section
7868 at the front about calling Exim under different names.
7870 51. Added "extra_headers" argument to the "mail" command in filter files.
7872 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7875 53. Installed eximstats 1.29.
7877 54. Added transport_filter_timeout as a generic transport option.
7879 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7880 Cc: header lines. This was required by RFC 822, but it not required by RFC
7883 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7884 incoming messages that did not have them. Now it does so only if the
7885 message originates locally, that is, if there is no associated remote host
7886 address. When Resent- header lines are present, this applies to the Resent-
7887 lines rather than the non-Resent- lines.
7889 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7890 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7892 58. Messages for configuration errors now include the name of the main
7893 configuration file - useful now that there may be more than one file in a
7894 list (.included file names were always shown).
7896 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7897 for those rare installations that do not start the daemon as root or run it
7898 setuid root. I've cut out the call to initgroups() if the daemon is not
7901 60. The Exim user and group can now be bound into the binary as text strings
7902 that are looked up at the start of Exim's processing.
7904 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7906 62. Added $mailstore_basename variable.
7908 63. Installed patch to sieve.c from Michael Haardt.
7910 64. When Exim failed to open the panic log after failing to open the main log,
7911 the original message it was trying to log was written to stderr and debug
7912 output, but if they were not available (the usual case in production), it
7913 was lost. Now it is written to syslog before the two lines that record the
7914 failures to open the logs.
7916 65. Users' Exim filters run in subprocesses under the user's uid. It is
7917 possible for a "deliver" command or an alias in a "personal" command to
7918 provoke an address rewrite. If logging of address rewriting is configured,
7919 this fails because the process is not running as root or exim. There may be
7920 a better way of dealing with this, but for the moment (because 4.30 needs
7921 to be released), I have disabled address rewrite logging when running a
7922 filter in a non-root, non-exim process.
7928 1. The buildconfig auxiliary program wasn't quoting the value set for
7929 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7930 not defined. This bug was masked in 4.22 by the effect that was fixed in
7933 2. Some messages that were rejected after a message id was allocated were
7934 shown as "incomplete" by exigrep. It no longer does this for messages that
7935 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7937 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7938 have allow_domain_literals set, the ID did not get logged in the <= line.
7939 Domain literals are now always recognized in Message-ID: header lines.
7941 4. The first argument for a ${extract expansion item is the key name or field
7942 number. Leading and trailing spaces in this item were not being ignored,
7943 causing some misleading effects.
7945 5. When deliver_drop_privilege was set, single queue runner processes started
7946 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7947 same command in the process it spins off) were not dropping privilege.
7949 6. When the daemon running as "exim" started a queue runner, it always
7950 re-executed Exim in the spun-off process. This is a waste of effort when
7951 deliver_drop_privilege is set. The new process now just calls the
7952 queue-runner function directly.
7958 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7961 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7962 ignored. Though the use of -oP was forcing the writing of a pid file, it
7963 was always written to the default place.
7965 3. If the message "no IP address found for host xxxx" is generated during
7966 incoming verification, it is now followed by identification of the incoming
7967 connection (so you can more easily find what provoked it).
7969 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7971 5. Added some features to "harden" Exim a bit more against certain attacks:
7973 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7974 be put in Local/Makefile. This is like the never_users runtime option,
7975 but it cannot be overridden. The default setting is "root".
7977 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7978 prefix string with which any file named in a -C command line option
7981 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7982 is retained for -C and -D only if the caller of Exim is root. Without
7983 it, the exim user may also use -C and -D and retain privilege.
7985 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7986 command line option is disabled.
7988 6. Macro names set by the -D option must start with an upper case letter, just
7989 like macro names defined in the configuration file.
7991 7. Added "dereference=" facility to LDAP.
7993 8. Two instances of the typo "uknown" in the source files are fixed.
7995 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7996 the Configure-Makefile script screwed up while processing it.
7998 10. Incorporated PCRE 4.4.
8000 11. The SMTP synchronization check was not operating right at the start of an
8001 SMTP session. For example, it could not catch a HELO sent before the client
8002 waited for the greeting. There is now a check for outstanding input at the
8003 point when the greeting is written. Because of the duplex, asynchronous
8004 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8005 way, but not yet received, when the check is performed.
8007 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8008 on TCP/IP sockets, because this apparently causes some broken clients to
8011 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8012 unchanged) from the Cygwin maintainer.
8014 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8015 was supported instead of testing for mbx. Effectively a typo.
8017 15. The spa authenticator server code was not checking that the input it
8018 received was valid base64.
8020 16. The debug output line for the "set" modifier in ACLs was not showing the
8021 name of the variable that was being set.
8023 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8025 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8026 lookup on its own. Something else was needed to trigger the lookup. For
8027 example, a match in host_lookup or the need for a host name in a host list.
8028 Now, if $sender_host_name is referenced and the host name has not yet been
8029 looked up, a lookup is performed. If the lookup fails, the variable remains
8030 empty, and $host_lookup_failed is set to "1".
8032 19. Added "eqi" as a case-independent comparison operator.
8034 20. The saslauthd authentication condition could segfault if neither service
8035 nor realm was specified.
8037 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8038 error message that was logged was misleading, and incoming SMTP
8039 connections were dropped. The message is now more accurate, and temporary
8040 errors are given to SMTP connections.
8042 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8043 (except RSET) with a 421 error, until QUIT is received. However, it was
8044 failing to send a response to QUIT.
8046 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8047 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8048 ACL is now run after the helo_try_verify_hosts code.
8050 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8051 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8052 case-independent, so other case variants are also recognized). Apparently
8053 some systems use these upper case variants.
8055 25. If more than two messages were waiting for the same host, and a transport
8056 filter was specified for the transport, Exim sent two messages over the
8057 same TCP/IP connection, and then failed with "socket operation on non-
8058 socket" when it tried to send the third.
8060 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8062 27. The extern definition of crypt16() in expand.c was not being excluded when
8063 the OS had its own crypt16() function.
8065 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8066 as a preferred synonym for return_size_limit, both as an option and as an
8069 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8071 30. Changed the default configuration ACL to relax the local part checking rule
8072 for addresses that are not in any local domains. For these addresses,
8073 slashes and pipe symbols are allowed within local parts, but the sequence
8074 /../ is explicitly forbidden.
8076 31. SPA server authentication was not clearing the challenge buffer before
8079 32. log_message in a "warn" ACL statement was writing to the reject log as
8080 well as to the main log, which contradicts the documentation and doesn't
8081 seem right (because no rejection is happening). So I have stopped it.
8083 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8084 However, I am unable to do any testing of this.
8086 34. Fixed an infelicity in the appendfile transport. When checking directories
8087 for a mailbox, to see if any needed to be created, it was accidentally
8088 using path names with one or more superfluous leading slashes; tracing
8089 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8091 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8092 contents are added to the log line that is written for every discarded
8093 recipient. (Previously a log_message setting was ignored.)
8095 36. The ${quote: operator now quotes the string if it is empty.
8097 37. The install script runs exim in order to find its version number. If for
8098 some reason other than non-existence or emptiness, which it checks, it
8099 could not run './exim', it was installing it with an empty version number,
8100 i.e. as "exim-". This error state is now caught, and the installation is
8103 38. An argument was missing from the function that creates an error message
8104 when Exim fails to connect to the socket for saslauthd authentication.
8105 This could cause Exim to crash, or give a corrupted message.
8107 39. Added isip, isip4, and isip6 to ${if conditions.
8109 40. The ACL variables $acl_xx are now saved with the message, and can be
8110 accessed later in routers, transports, and filters.
8112 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8113 strings in the file are not string-expanded.
8115 42. If a MAIL command specified a SIZE value that was too large to fit into an
8116 int variable, the check against message_size_limit failed. Such values are
8117 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8118 day this will have to be increased, but I don't think I want to be around
8119 when emails are that large.
8126 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8127 iconv() is not standard in FreeBSD.
8129 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8130 IPv6 enabled. The observed symptom was a segmentation fault on return from
8131 the function os_common_find_running_interfaces() in src/os.c.
8133 3. In the check_special_case() function in daemon.c I had used "errno" as an
8134 argument name, which causes warnings on some systems. This was basically a
8135 typo, since it was named "eno" in the comments!
8137 4. The code that waits for the clock to tick (at a resolution of some fraction
8138 of a second) so as to ensure message-id uniqueness was always waiting for
8139 at least one whole tick, when it could have waited for less. [This is
8140 almost certainly not relevant at current processor speeds, where it is
8141 unlikely to ever wait at all. But we try to future-proof.]
8143 5. The function that sleeps for a time interval that includes fractions of a
8144 second contained a race. It did not block SIGALRM between setting the
8145 timer, and suspending (a couple of lines later). If the interval was short
8146 and the sigsuspend() was delayed until after it had expired, the suspension
8147 never ended. On busy systems this could lead to processes getting stuck for
8150 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8151 process, before it forks any delivery processes. The open lookup caching
8152 mechanism meant that the open file or database connection was passed into
8153 the delivery process. The problem was that delivery processes always tidy
8154 up cached lookup data. This could cause a problem for the next delivery
8155 process started by the queue runner, because the external queue runner
8156 process does not know about the closure. So the next delivery process
8157 still has data in the lookup cache. In the case of a file lookup, there was
8158 no problem because closing a file descriptor in a subprocess doesn't affect
8159 the parent. However, if the lookup was caching a connection to a database,
8160 the connection was closed, and the second delivery process was likely to
8161 see errors such as "PGSQL: query failed: server closed the connection
8162 unexpectedly". The problem has been fixed by closing all cached lookups
8163 in a queue runner before running a delivery process.
8165 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8166 seem to have the "const" qualifier which it has on other OS. I've
8169 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8170 *of the same type* (client or server) with the same public name that an
8171 error should be diagnosed.
8173 9. When Exim looked up a host name for an IP address, but failed to find the
8174 original IP address when looking up the host name (a safety check), it
8175 output the message "<ip address> does not match any IP for NULL", which was
8176 confusing, to say the least. The bug was that the host name should have
8177 appeared instead of "NULL".
8179 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8180 user that is built into the binary, and the -C or -D options is used, root
8181 privilege is dropped before the configuration file is read. In addition,
8182 logging is switched to stderr instead of the normal log files. If the
8183 configuration then re-defines the Exim user, the unprivileged environment
8184 is probably not what is expected, so Exim logs a panic warning message (but
8187 However, if deliver_drop_privilege is set, the unprivileged state may well
8188 be exactly what is intended, so the warning has been cut out in that case,
8189 and Exim is allowed to try to write to its normal log files.
8195 1. smtp_return_error_details was not giving details for temporary sender
8196 or receiver verification errors.
8198 2. Diagnose a configuration error if two authenticators have the same public
8201 3. Exim used not to create the message log file for a message until the first
8202 delivery attempt. This could be confusing when incoming messages were held
8203 for policy or load reasons. The message log file is now created at the time
8204 the message is received, and an initial "Received" line is written to it.
8206 4. The automatically generated man page for command line options had a minor
8207 bug that caused no ill effects; however, a more serious problem was that
8208 the procedure for building the man page automatically didn't always
8209 operate. Consequently, release 4.20 contains an out-of-date version. This
8210 shouldn't happen again.
8212 5. When building Exim with embedded Perl support, the script that builds the
8213 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8214 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8216 6. The freeze_tell option was not being used for messages that were frozen on
8217 arrival, either by an ACL or by local_scan().
8219 7. Added the smtp_incomplete_transaction log selector.
8221 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8222 was accepting AUTH without a new EHLO.
8224 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8225 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8228 10. Logging of TCP/IP connections (when configured) now happens in the main
8229 daemon process instead of the child process, so that the TCP/IP connection
8230 count is more accurate (but it can never be perfect).
8232 11. The use of "drop" in a nested ACL was not being handled correctly in the
8233 outer ACL. Now, if condition failure induced by the nested "drop" causes
8234 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8235 or "require"), the connection is dropped.
8237 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8238 that yield "discard" can now be used with an "accept" or a "discard" verb,
8239 but an error is generated for any others (because I can't see a useful way
8240 to define what should happen).
8242 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8243 are now processed in the same way as lines in the Exim configuration file.
8244 In particular, continuation lines are supported.
8246 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8248 15. Added -ti meaning -t -i.
8250 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8251 domains, and warn by logging if others are found.
8253 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8254 returned from GIFCONF when Exim is trying to find the list of interfaces on
8255 a host. The code in os.c has been modified to copy each ifreq to an aligned
8256 structure in all cases.
8258 Also, in some cases, the returned ifreq's were being copied to a 'struct
8259 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8260 means the last couple of bytes of an IPv6 address could be chopped if the
8261 ifreq contained only a normal sockaddr (14 bytes storage).
8263 18. Named domain lists were not supported in the hosts_treat_as_local option.
8264 An entry such as +xxxx was not recognized, and was treated as a literal
8267 19. Ensure that header lines added by a DATA ACL are included in the reject log
8268 if the ACL subsequently rejects the message.
8270 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8271 MD5 (which is deprecated).
8273 21. When testing a filter file using -bf, Exim was writing a message when it
8274 took the sender from a "From " line in the message, but it was not doing so
8275 when it took $return_path from a Return-Path: header line. It now does.
8277 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8278 with a valid header line field name (a series of printing characters
8279 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8281 23. Changed "disc" in the source to "disk" to conform to the documentation and
8282 the book and for uniformity.
8284 24. Ignore Sendmail's -Ooption=value command line item.
8286 25. When execve() failed while trying to run a command in a pipe transport,
8287 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8288 could be confused with a return value of 69 from the command itself. This
8289 has been changed to 127, the value the shell returns if it is asked to run
8290 a non-existent command. The wording for the related log line suggests a
8291 non-existent command as the problem.
8293 26. If received_header_text expands to an empty string, do not add a Received:
8294 header line to the message. (Well, it adds a token one on the spool, but
8295 marks it "old" so that it doesn't get used or transmitted.)
8297 27. Installed eximstats 1.28 (addition of -nt option).
8299 28. There was no check for failure on the call to getsockname() in the daemon
8300 code. This can fail if there is a shortage of resources on the system, with
8301 ENOMEM, for example. A temporary error is now given on failure.
8303 29. Contrary to the C standard, it seems that in some environments, the
8304 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8305 program. Exim now does this explicitly; it affects the formatting of
8306 timestamps using strftime().
8308 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8309 complaints. I've now initialized all the variables, to avoid this.
8311 32. Header lines added by a system filter were not being "seen" during
8312 transport-time rewrites.
8314 33. The info_callback() function passed to OpenSSL is set up with type void
8315 (*)(SSL *, int, int), as described somewhere. However, when calling the
8316 function (actually a macro) that sets it up, the type void(*)() is
8317 expected. I've put in a cast to prevent warnings from picky compilers.
8319 34. If a DNS black list lookup found a CNAME record, but there were no A
8320 records associated with the domain it pointed at, Exim crashed.
8322 35. If a DNS black list lookup returned more than one A record, Exim ignored
8323 all but the first. It now scans all returned addresses if a particular IP
8324 value is being sought. In this situation, the contents of the
8325 $dnslist_value variable are a list of all the addresses, separated by a
8328 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8329 to accept a host name and all its aliases if the forward lookup for any of
8330 them yielded the IP address of the incoming connection. Now it accepts only
8331 those names whose forward lookup yields the correct IP address. Any other
8332 names are discarded. This closes a loophole whereby a rogue DNS
8333 administrator could create reverse DNS records to break through a
8334 wildcarded host restriction in an ACL.
8336 37. If a user filter or a system filter that ran in a subprocess used any of
8337 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8338 the wrong values were passed to the pipe command ($thisaddress had the
8339 value of $0, $0 had the value of $1, etc). This bug was introduced by
8340 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8342 38. Improved the line breaking for long SMTP error messages from ACLs.
8343 Previously, if there was no break point between 40 and 75 characters, Exim
8344 left the rest of the message alone. Two changes have been made: (a) I've
8345 reduced the minimum length to 35 characters; (b) if it can't find a break
8346 point between 35 and 75 characters, it looks ahead and uses the first one
8347 that it finds. This may give the occasional overlong line, but at least the
8348 remaining text gets split now.
8350 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8351 file descriptors might be low, and that setting 1000 would always raise it.
8352 It turns out that in some environments, the limit is already over 1000 and
8353 that lowering it causes trouble. So now Exim takes care not to decrease it.
8355 40. When delivering a message, the value of $return_path is set to $sender_
8356 address at the start of routing (routers may change the value). By an
8357 oversight, this default was not being set up when an address was tested by
8358 -bt or -bv, which affected the outcome if any router or filter referred to
8361 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8362 log only when "message" or "log_message" are set. However, if one of the
8363 conditions was an address verification, or a call to a nested ACL, the
8364 messages generated by the underlying test were being passed through. This
8365 no longer happens. The underlying message is available in $acl_verify_
8366 message for both "message" and "log_message" expansions, so it can be
8367 passed through if needed.
8369 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8370 new expansion $bh_ to give the encoded byte string without charset
8371 translation. Translation happens only if iconv() is available; HAVE_ICONV
8372 indicates this at build time. HEADERS_CHARSET gives the charset to
8373 translate to; headers_charset can change it in the configuration, and
8374 "headers charset" can change it in an individual filter file.
8376 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8377 that creates RFC 2047 encoded "words" labels them as that charset instead
8378 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8379 expansion operator; (ii) when Exim creates a From: line for a local
8380 message; (iii) when a header line is rewritten to include a "phrase" part.
8382 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8383 buggy, causing it to skip the first lines of messages whose message ID
8384 ended in 'D'. This would not have bitten before Exim release 4.14, because
8385 message IDs were unlikely to end in 'D' before then. The effect was to have
8386 incorrect size information for certain domains.
8388 45. #include "config.h" was missing at the start of the crypt16.c module. This
8389 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8392 46. If there was a timeout during a "random" callout check, Exim treated it as
8393 a failure of the random address, and carried on sending RSET and the real
8394 address. If the delay was just some slowness somewhere, the response to the
8395 original RCPT would be taken as a response to RSET and so on, causing
8396 mayhem of various kinds.
8398 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8399 when I implemented it. It didn't allow for the fact that some option values
8400 may legitimately be negative (e.g. size_addition), and it didn't even do
8401 the right test for positive values.
8403 48. Domain names in DNS records are case-independent. Exim always looks them up
8404 in lower case. Some resolvers return domain names in exactly the case they
8405 appear in the zone file, that is, they may contain uppercase letters. Not
8406 all resolvers do this - some return always lower case. Exim was treating a
8407 change of case by a resolver as a change of domain, similar to a widening
8408 of a domain abbreviation. This triggered its re-routing code and so it was
8409 trying to route what was effectively the same domain again. This normally
8410 caused routing to fail (because the router wouldn't handle the domain
8411 twice). Now Exim checks for this case specially, and just changes the
8412 casing of the domain that it ultimately uses when it transmits the message
8415 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8418 50. If a filter generated a file delivery with a non-absolute name (possible if
8419 no home directory exists for the router), the forbid_file option was not
8422 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8423 the existing equality matching.
8425 52. Exim was using ints instead of ino_t variables in some places where it was
8426 dealing with inode numbers.
8428 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8429 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8430 called TMPDIR, and if it finds it is different, it changes its value.
8432 54. The smtp_printf() function is now made available to local_scan() so
8433 additional output lines can be written before returning. There is also an
8434 smtp_fflush() function to enable the detection of a dropped connection.
8435 The variables smtp_input and smtp_batched_input are exported to
8438 55. Changed the default runtime configuration: the message "Unknown user"
8439 has been removed from the ACL, and instead placed on the localuser router,
8440 using the cannot_route_message feature. This means that any verification
8441 failures that generate their own messages won't get overridden. Similarly,
8442 the "Unrouteable address" message that was in the ACL for unverifiable
8443 relay addresses has also been removed.
8445 56. Added hosts_avoid_esmtp to the smtp transport.
8447 57. The exicyclog script was not checking for the esoteric option
8448 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8449 will work only if exicyclog is run under the appropriate euid.
8451 58. Following a discussion on the list, the rules by which Exim recognises line
8452 endings on incoming messages have been changed. The -dropcr and drop_cr
8453 options are now no-ops, retained only for backwards compatibility. The
8454 following line terminators are recognized: LF CRLF CR. However, special
8455 processing applies to CR:
8457 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8458 nor a local message in the state where . is a terminator.
8460 (ii) If a bare CR is encountered in a header line, an extra space is added
8461 after the line terminator so as not to end the header. The reasoning
8462 behind this is that bare CRs in header lines are most likely either
8463 to be mistakes, or people trying to play silly games.
8465 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8466 was being incorrectly given as 18 bytes larger than it should have been.
8467 This is a VOB (very old bug).
8469 60. This may never have affected anything current, but just in case it has:
8470 When the local host is found other than at the start of a list of hosts,
8471 the local host, those with the same MX, and any that follow, are discarded.
8472 When the list in question was part of a longer list of hosts, the following
8473 hosts (not currently being processed) were also being discarded. This no
8474 longer happens. I'm not sure if this situation could ever has previously
8477 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8480 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8481 header. This is something that is recommended in a new Internet Draft, and
8482 is something that is documented as being done by Sendmail. There are two
8483 possible values. For messages generated by the autoreply transport, Exim
8486 Auto-Submitted: auto-replied
8488 whereas for all other generated messages (e.g. bounces) it adds
8490 Auto-Submitted: auto-generated
8492 63. The "personal" condition in filters now includes a test for the
8493 Auto-Submitted: header. If it contains the string "auto-" the message it
8494 not considered personal.
8496 64. Added rcpt_include_affixes as a generic transport option.
8498 65. Added queue_only_override (default true).
8500 66. Added the syslog_duplication option.
8502 67. If what should have been the first header line of a message consisted of
8503 a space followed by a colon, Exim was mis-interpreting it as a header line.
8504 It isn't of course - it is syntactically invalid and should therefore be
8505 treated as the start of the message body. The misbehaviour could have
8506 caused a number of strange effects, including loss of data in subsequent
8507 header lines, and spool format errors.
8509 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8510 client host had authenticated. This control can now be exercised by an ACL
8511 for more flexibility.
8513 69. By default, callouts do not happen when testing with -bh. There is now a
8514 variant, -bhc, which does actually run the callout code, including
8515 consulting and updating the callout cache.
8517 70. Added support for saslauthd authentication, courtesy of Alexander
8520 71. If statvfs() failed on the spool or log directories while checking their
8521 size for availability, Exim confusingly gave the error "space shortage".
8522 Furthermore, in debugging mode it crashed with a floating point exception.
8523 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8524 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8525 serious problem, Exim now writes to the main and panic logs when this
8526 happens, with details of the failure. It then refuses to accept the
8527 incoming message, giving the message "spool directory problem" or "log
8528 directory problem" with a 421 code for SMTP messages.
8530 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8531 0, 1, and 2 exist, because some OS complain for execs without them (see
8532 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8533 descriptors. However, the code omitted to check that the open succeeded,
8534 causing mysterious errors if for some reason the permissions on /dev/null
8535 got screwed. Now Exim writes a message to the main and panic logs, and
8536 bombs out if it can't open /dev/null.
8538 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8539 interact so that it is all more flexible. It is supposed to remain
8540 backwards compatible. Also added extra_local_interfaces.
8542 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8543 to bomb out with an assertion failure - to the client this appears as a
8544 connection drop. This problem occurs in the part of the code that was taken
8545 from the Samba project. Fortunately, the assertion is in a very simple
8546 function, so I have fixed this by reproducing the function inline in the
8547 one place where it is called, and arranging for authentication to fail
8548 instead of killing the process with assert().
8550 75. The SPA client code was not working when the server requested OEM rather
8551 than Unicode encoding.
8553 76. Added code to make require_files with a specific uid setting more usable in
8554 the case where statting the file as root fails - usually a non-root-mounted
8555 NFS file system. When this happens and the failure is EACCES, Exim now
8556 forks a subprocess and does the per-uid checking as the relevant uid.
8558 77. Added process_log_path.
8560 78. If log_file_path was not explicitly set, a setting of check_log_space or
8561 check_log_inodes was ignored.
8563 79. If a space check for the spool or log partitions fails, the incident is now
8564 logged. Of course, in the latter case the data may get lost...
8566 80. Added the %p formatting code to string_format() so that it can be used to
8567 print addresses in debug_print(). Adjusted all the address printing in the
8568 debugging in store.c to use %p rather than %d.
8570 81. There was a concern that a line of code in smtp_in.c could overflow a
8571 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8572 initially expressed, the concern was not well-founded, because trailing
8573 spaces are removed early. However, if the trailing spaces were followed by
8574 a NULL, they did not get removed, so the overflow was possible. Two fixes
8577 (a) I re-wrote the offending code in a cleaner fashion.
8578 (b) If an incoming SMTP command contains a NULL character, it is rejected
8581 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8582 runs initgroups(), so that if the Exim user is in any additional groups,
8583 they will be used during message reception.
8589 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob