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.
107 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
108 SMTP connection" log lines.
110 JH/02 Option default value updates:
111 - queue_fast_ramp (main) true (was false)
112 - remote_max_parallel (main) 4 (was 2)
114 JH/03 Cache static regex pattern compilations, for use by ACLs.
116 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
117 Make the rewrite never match and keep the logging. Trust the
118 admin to be using verify=header-syntax (to actually reject the message).
120 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
121 (under Linux) we watch the dir containing the final file; previously
122 it would be the dir with the first symlink. We still do not monitor
125 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
126 dn_expand() is happy to pass them through.
128 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
129 the file with the proof had an unchanged name, the new proof(s) were
130 loaded on top of the old ones (and nover used; the old ones were stapled).
132 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
133 more than one message arrived in a single connection a reference from
134 the earlier message could be re-used. Often a sigsegv resulted.
135 These variables were introduced in Exim 4.87.
136 Debug help from Graeme Fowler.
138 JH/09 Fix ${filter } for conditions that modify $value. Previously the
139 modified version would be used in construction the result, and a memory
142 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
143 Find and fix by Jasen Betts.
145 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
146 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
147 the systemwide configuration to override the Exim config.
149 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
150 API changes in libopendmarc.
152 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
153 pid 1, in the normal "background daemon" mode, having to drop process-
154 group leadership also lost track of needing to create listener sockets.
156 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
157 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
159 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
160 a capture group which obtained no text (eg. "(abc)*" matching zero
161 occurrences) could cause a segfault if the corresponding $<n> was
164 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
165 included a close-brace character (eg. it itself used an expansion) an
168 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
169 starting TLS. Previously it was after, meaning that attackers on such
170 ports had to be screened using the host_reject_connection main config
171 option. The new sequence aligns better with the STARTTLS behaviour, and
172 permits defences against crypto-processing load attacks, even though it
173 is strictly an incompatible change.
174 Also, avoid sending any SMTP fail response for either the connect ACL
175 or host_reject_connection, for TLS-on-connect ports.
177 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
178 Previously this was not permitted, but it makes reasonable sense.
179 While there, restore a restriction on using it from a connect ACL; given
180 the change JH/16 it could only return false (and before 4.91 was not
183 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
184 was exactly sized compared to the log buffer, a crash occurred with the
185 misleading message "bad memory reference; pool not found".
186 Found and traced by Jasen Betts.
188 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
189 dns_again_means_nonexist included an element causing a DNS lookup which
190 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
191 included (though probably not limited to) a process crash from stack
192 memory limit, or from excessive open files. Replace this with a paniclog
193 whine (as this is likely a configuration error), and returning
196 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
197 this always failed, probably leading to the usual downgrade to in-clear
200 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
201 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
202 to not checking that list for these lookups.
204 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
207 JH/23 Fix crash in string expansions. Previously, if an empty variable was
208 immediately followed by an expansion operator, a null-indirection read
209 was done, killing the process.
211 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
212 include an SMTP response string which is longer than that supported
213 by the delivering transport. Alleviate by wrapping such lines before
216 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
217 chars (RFC limit). Previously a limit of 12 items was made, which with
218 a not-impossible References: in the message being bounced could still
219 be over-large and get stopped in the transport.
221 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
222 close. Previously a bare socket close was done.
224 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
227 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
228 message being considered for delivery was already being handled by
229 another process, and having an SMTP connection already open, the function
230 to close it tried to use an uninitialized variable. This would afftect
231 high-volume sites more, especially when running mailing-list-style loads.
232 Pollution of logs was the major effect, as the other process delivered
233 the message. Found and partly investigated by Graeme Fowler.
235 JH/29 Change format of the internal ID used for message identification. The old
236 version only supported 31 bits for a PID element; the new 64 (on systems
237 which can use Base-62 encoding, which is all currently supported ones
238 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
239 and must use Base-36). The new ID is 23 characters rather than 16, and is
240 visible in various places - notably logs, message headers, and spool file
241 names. Various of the ancillary utilities also have to know the format.
242 As well as the expanded PID portion, the sub-second part of the time
243 recorded in the ID is expanded to support finer precision. Theoretically
244 this permits a receive rate from a single comms channel of better than the
246 The major timestamp part of the ID is not changed; at 6 characters it is
247 usable until about year 3700.
248 Updating from previously releases is fully supported: old-format spool
249 files are still usable, and the utilities support both formats. New
250 message will use the new format. The one hints-DB file type which uses
251 message-IDs (the transport wait- DB) will be discarded if an old-format ID
252 is seen; new ones will be built with only new-format IDs.
253 Optionally, a utility can be used to convert spool files from old to new,
254 but this is only an efficiency measure not a requirement for operation
255 Downgrading from new to old requires running a provided utility, having
256 first stopped all operations. This will convert any spool files from new
257 back to old (losing time-precision and PID information) and remove any
258 wait- hints databases.
260 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
261 we treated them as item separators when parsing for a list item, but they
262 need to be protected by the doublequotes. While there, add handling for
265 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
266 Found and fixed by Jasen Betts. No testcase for this as my usual text
267 editor insists on emitting only valid UTF-8.
269 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
270 a null-indirection SIGSEGV for the receive process.
272 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
273 -bh use. Internal checking would spot this and take a panic.
275 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
276 In 4.96 this would expand to empty.
278 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
279 certificate. Find and fix by Andreas Metzler.
281 JH/36 Add ARC info to DMARC hostory records.
283 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
284 or fakedefer. Previously the sender could discover that the message
285 had in fact been accepted.
287 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
288 sequences. Previously the input was not noted as being tainted; notably
289 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
290 bad coding of authenticators.
292 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
293 and ${tr...}. Found and diagnosed by Heiko Schlichting.
295 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
296 could be triggered by externally-supplied input. Found by Trend Micro.
299 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
300 be triggered by externally-controlled input. Found by Trend Micro.
303 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
304 be triggered by externally-controlled input. Found by Trend Micro.
307 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
308 Make the rewrite never match and keep the logging. Trust the
309 admin to be using verify=header-syntax (to actually reject the message).
311 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
314 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
320 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
321 after reception to before a subsequent reception. This should
322 mean slightly faster delivery, and also confirmation of reception
325 JH/02 Move from using the pcre library to pcre2. The former is no longer
326 being developed or supported (by the original developer).
328 JH/03 Constification work in the filters module required a major version
329 bump for the local-scan API. Specifically, the "headers_charset"
330 global which is visible via the API is now const and may therefore
331 not be modified by local-scan code.
333 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
334 sendfile() didi not account for the way the ClamAV driver code called it.
336 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
337 time check was being done for every character; replace that with one
340 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
341 sent was prefixed with a length byte.
343 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
344 RFC 5321. Previously Dovecot (at least) would log errors during
347 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
348 in 4.61 without a reason logged. Binaries will be bigger, which might
349 matter on diskspace-constrained systems, but debug is easier.
351 JH/09 Fix macro-definition during "-be" expansion testing. The move to
352 write-protected store for macros had not accounted for these runtime
353 additions; fix by removing this protection for "-be" mode.
355 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
356 handing out large-numbered file descriptors, violating the usual Unix
357 assumption (and required by Posix) that the lowest possible number will be
358 allocated by the kernel when a new one is needed. In the daemon, and any
359 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
360 are not useable for FD_SET() [and hence select()] and overwrite the stack.
361 Assorted crashes happen.
363 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
364 main-section options or in a connect ACL, the value from the first ever
365 connection was never replaced for subsequent connections. Found by
368 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
369 though only once PCRE2 was introduced: the memory accounting used under
370 debug offset allocations by an int, giving a hard trap in early startup.
371 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
373 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
374 with underbars is given. The write-protection of configuration introduced
375 in 4.95 trapped when normalisation was applied to an option not needing
378 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
380 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
381 failing to close the certificates directory, every hour or any time it
384 JH/16 Debugging initiated by an ACL control now continues through into routing
385 and transport processes. Previously debugging stopped any time Exim
386 re-execs, or for processing a queued message.
388 JH/17 The "expand" debug selector now gives more detail, specifically on the
389 result of expansion operators and items.
391 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
392 bad comparison between the option value and the name of the file to
393 be included was done, and a mismatch was wrongly identified.
394 4.88 to 4.95 are affected.
396 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
398 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
399 supplied as the db file-pair basename. Previously, if a directory
400 path was given, for example via the autoreply "once" option, the DB
401 file.pag and file.dir files would be created in that directory's
404 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
405 "taint" log_selector. These were previously deprecated.
407 JH/22 Fix static address-list lookups to properly return the matched item.
408 Previously only the domain part was returned.
410 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
411 the call into OpenSSL to send a TLS Close was being repeated; this
412 resulted in the library waiting for the peer's Close. If that was never
413 sent we waited forever. Fix by tracking send calls.
415 JH/24 The ${run} expansion item now expands its command string elements after
416 splitting. Previously it was before; the new ordering makes handling
417 zero-length arguments simpler. The old ordering can be obtained by
418 appending a new option "preexpand", after a comma, to the "run".
420 JH/25 Taint-check exec arguments for transport-initiated external processes.
421 Previously, tainted values could be used. This affects "pipe", "lmtp" and
422 "queryprogram" transport, transport-filter, and ETRN commands.
423 The ${run} expansion is also affected: in "preexpand" mode no part of
424 the command line may be tainted, in default mode the executable name
427 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
428 the the facility was not passed across execs, and only the first message
429 passed over a connection could use BDAT; any further ones using DATA.
431 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
432 uses $sending_ip_address and an interface is specified.
433 Previously any use of the local address in the EHLO name disabled
434 PIPECONNECT, the common case being to use the rDNS of it.
436 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
437 resumption. Previously verify failed because no certificate status is
438 passed on the wire for the restarted session. Fix by using the recorded
439 ocsp status of the stored session for the new connection.
441 JH/29 TLS resumption: the key for session lookup in the client now includes
442 more info that a server could potentially use in configuring a TLS
443 session, avoiding oferring mismatching sessions to such a server.
444 Previously only the server IP was used.
446 JH/30 Fix string_copyn() for limit greater than actual string length.
447 Previously the copied amount was the limit, which could result in a
448 overlapping memcpy for newly allocated destination soon after a
449 source string shorter than the limit. Found/investigated by KM.
451 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
452 close; it may be needed for a subsequent connection. This caused a
453 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
455 JH/32 Fix CHUNKING for a second message on a connection when the first was
456 rejected. Previously we did not reset the chunking-offered state, and
457 erroneously rejected the BDAT command. Investigation help from
460 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
461 an empty address. Previously the expansion returned an error.
463 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
464 proxy. Previously these were misparsed, leading to paniclog entries.
470 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
471 related applications. Previously an "H" was used where available info
472 says that "M" should be, so change to match.
474 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
475 as arguments, so an implementation trying to copy these into a local
476 buffer was taking a taint-enforcement trap. Fix by using dynamically
477 created buffers. Similar fix for radius expansion condition.
479 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
480 reasonable, eg. to count headers. Fix by using dynamically created
481 buffers rather than a local. Do similar fixes for ACL actions "dcc",
482 "log_reject_target", "malware" and "spam"; the arguments are expanded
483 so could be handling tainted values.
485 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
486 broken the (no-op) support for this sendmail command. Restore it
487 to doing nothing, silently, and returning good status.
489 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
490 record path was given (or the default used) without a leading directory
491 path, an error occurred on trying to open it. Use the transport's working
494 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
495 transport. Previously it was the name on the DNS A-record; use instead
496 the head of the CNAME chain leading there (if there is one). This seems
497 to align better with RFC 6125.
499 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
500 smtp_accept_max_per_host allocated resources which were not released
501 when the limit was exceeded. This eventually crashed the daemon. Fix
502 by adding a release action in that path.
504 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
505 expanded; previously using tainted values was rejected. Fix by using
506 dynamically-created buffers.
508 JH/09 Relax restrictions on ACL verify condition needing access to message
509 headers. Previously they were only permitted in data and non-smtp ACLs;
510 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
511 not_blind, header_sender and header_names_ascii verification.
513 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
514 Previously a macro used one argument twice; when called with the
515 argument as an expression having side-effects, incorrect operation
516 resulted. Use an inlineable function.
518 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
519 held open for a verify callout. Previously this wan not accounted for
520 and a corrupt onward SMTP conversation resulted.
522 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
523 Previously they were embedded naively in the constructed address; when
524 needed, strip the quoting and quote the entire local_part.
525 Also make the inbound_srs expansion condition handle quoting.
527 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
528 excluded, not matching the documentation.
530 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
531 was given for the sqlite_dbfile a trap resulted.
533 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
534 "name" argument resulted in a trap. There is no reason to disallow such;
535 this was a coding error.
537 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
538 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
539 spent suspended, ignoring the POSIX definition. Previously we assumed
540 it did and a constant offset from real time could be used as a correction.
541 Change to using the same clock source for the start-of-message and the
542 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
543 exists, just to get a clock slightly more aligned to reality.
545 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
546 RFC says it is optional some validators care. The missing char was not
547 intended but triggered by a line-wrap alignment. Discovery and fix by
548 Guillaume Outters, hacked on by JH.
550 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
551 name being quoted was tainted a trap would be taken. Fix by using
552 dynamically created buffers. The routine could have been called by a
553 rewrite with the "h" flag, by using the "-F" command-line option, or
554 by using a "name=" option on a control=submission ACL modifier.
556 JH/19 SPF: change the Authentication-Results expansion component to give
557 smtp.helo when the sender domain is empty. Previously it gave
560 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
561 not in 127.0.0.0/8 to help in spotting list domains taken over by a
562 domain-parking registrar.
564 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
565 Previously when a whitespace character was specified it was not inserted
566 after removing the newline.
568 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
569 the domain part of the recipient address. This overrides any tls_sni
570 option set, which was previously used.
572 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
575 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
576 is_tainted() had an off-by-one error in the overenthusiastic direction.
577 Find and fix by Gavan. Although NetBSD is not a supported platform for
578 4.94 this bug could affect other platforms.
580 PP/01 Fix default prime selection to be consistent.
581 One path used ike23 still, instead of exim.dev.20160529.3; now both
582 execution flows will use the same DH primes (currently
583 exim.dev.20160529.3).
585 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
586 option in smtp transport, to match the documentation. Previously
587 verification was not being done in some cases where it should have been.
589 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
590 than one server was defined and depending on the platform memory layout
591 details, an internal consistency trap could be hit while walking the list
594 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
595 files. The value is used by the authresults expansion item. Previously
596 if this was used in a router or transport, a crash could result.
598 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
599 have a sufficient incidence of EINTR returns from select that an
600 interaction with other operations done by the main daemon loop exposed
601 a bug in the error-handling. This was benign apart from the log
604 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
605 for consistency with delivered "=>" and failed "**" lines. While we're
606 there, handle PRX and TFO.
608 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
609 applied. This resulted, if any header-line rewrite rules were configured,
610 in a panic-log triggerable by sending a message with a long address in
611 a header. Fix by increasing the arbitrary limit to larger than a single
612 (dewrapped) 5322 header line maximum size.
614 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
615 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
616 RFC 6648 which deprecates X- options in protocols as a general practice.
617 Changeover between the implementations is handled by the mechanisms
620 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
621 a remote which uses callout/hold. Previously the local was queued.
623 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
626 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
627 log line. The asterisk is now consistently appended to the remote IP
628 (and port, if given), and will also be provided on defer and fail log
629 lines. Previously it could be placed on the local IP if that was being
630 logged, and was only provided on delivery lines.
632 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
634 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
635 authenticator client_send option. Previously the next char, after a pair
636 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
637 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
638 documentation. There is still no way to get a leading ^ immediately
639 after a NUL (ie. for the password of a PLAIN method authenticator.
641 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
642 files. For bad sizes read, delete the record and whine to paniclog.
644 JH/38 When logging an AUTH failure, as server, do not include sensitive
645 information. Previously, the credentials would be included if given
646 as part of the AUTH command line and an ACL denied authentication.
648 JH/39 Bug 2691: fix $local_part_data. When the matching list element
649 referred to a file, bad data was returned. This likely also affected
652 JH/40 The gsasl authenticator now supports caching of the salted password
653 generated by the client-side implementation. This required the addition
654 of a new variable: $auth4.
656 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
657 left undeleted; the attempt to re-create it then failed - resulting in
658 the usual "SIGHUP tp have daemon reload configuration" to not work.
659 This affected any platform not supporting "abstract" Unix-domain
660 sockets (i.e. not Linux).
662 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
663 recipients" response to RCPT in a later response, with a 250. The
664 previous coding assumed this would not happen, and under PIPELINING
665 would result in both lost and duplicate recipients for a message.
667 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
668 Previously the weighting was incorrectly applied. Similar fix for socks
669 proxies. Found and fixed by Heiko Schlichting.
671 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
672 not handle sub-lists included using the +namedlist syntax. While
673 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
674 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
677 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
678 Previously the permanent pool was used, so the sore could not be freed.
679 This meant a connection with many messages would use continually-growing
682 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
683 per-pool so as not to waste too much space. Previously a constant size
684 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
685 DOS attacks harder. The cost is wasted memory use in the larger blocks.
687 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
688 same space repeatedly, and a smaller process footprint.
690 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
691 for GnuTLS. Previously the name started "%s" which, while not a bug,
692 looked as if if might be one.
694 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
695 "continued connection") make the $proxy_* variables available. Previously
696 the information was not passed across the exec() call for subsequent
697 transport executions. This also mean that the log lines for the
698 messages can show the proxy information.
700 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
701 receive time. With modern systems the difference is significant.
702 The historical behaviour can be restored by disabling (a new) log_selector
703 "queue_time_exclusive".
705 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
706 content) were specifically tested for. Now, also cover expansions
707 resulting in ACL names and inline ACL content.
709 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
710 making it unusable in complex expressions.
712 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
713 fail with a taint-check on the spoolfile name, and leave the message
716 HS/01 Enforce absolute PID file path name.
718 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
720 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
722 PP/02 Bug 2643: Correct TLS DH constants.
723 A missing NUL termination in our code-generation tool had led to some
724 incorrect Diffie-Hellman constants in the Exim source.
725 Reported by kylon94, code-gen tool fix by Simon Arlott.
727 PP/03 Impose security length checks on various command-line options.
728 Fixes CVE-2020-SPRSS reported by Qualys.
730 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
731 better. Reported by Qualys.
733 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
734 providing a particularly obnoxious sender full name.
737 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
739 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
742 PP/08 Change default for recipients_max from unlimited to 50,000.
744 PP/09 Fix security issue with too many recipients on a message (to remove a
745 known security problem if someone does set recipients_max to unlimited,
746 or if local additions add to the recipient list).
747 Fixes CVE-2020-RCPTL reported by Qualys.
749 PP/10 Fix security issue in SMTP verb option parsing
750 Fixes CVE-2020-EXOPT reported by Qualys.
752 PP/11 Fix security issue in BDAT state confusion.
753 Ensure we reset known-good where we know we need to not be reading BDAT
754 data, as a general case fix, and move the places where we switch to BDAT
755 mode until after various protocol state checks.
756 Fixes CVE-2020-BDATA reported by Qualys.
758 HS/03 Die on "/../" in msglog file names
760 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
761 the Exim runtime user are allowed to create files.
763 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
766 QS/03 When reading the output from interpreted forward files we do not
767 pass the pipe between the parent and the interpreting process to
768 executed child processes (if any).
770 QS/04 Always die if requested from internal logging, even is logging is
773 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
774 the API; compilation noo longer completes with DMARC support included.
775 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
776 been reported on other platforms.
778 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
780 JH/56 Make the majority of info read from config files readonly, for defence-in-
781 depth against exploits. Suggestion by Qualys.
782 Not supported on Solaris 10.
784 JH/57 Fix control=fakreject for a custom message containing tainted data.
785 Previously this resulted in a log complaint, due to a re-expansion present
786 since fakereject was originally introduced.
788 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
789 modifier was given, a loop resulted.
791 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
792 pipelined SMTP command followed the BDAT LAST then it would be
793 incorrectly treated as part of the message body, causing a verification
796 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
797 require looser validation rules than those for 821-level addresses,
798 which only permit IP addresses.
804 JH/01 Avoid costly startup code when not strictly needed. This reduces time
805 for some exim process initialisations. It does mean that the logging
806 of TLS configuration problems is only done for the daemon startup.
808 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
810 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
811 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
814 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
815 DKIM signing is being done. Previously a transport_filter would always
816 disable CHUNKING, falling back to traditional DATA.
818 JH/05 Regard command-line recipients as tainted.
820 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
822 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
823 PAM library frees one of the arguments given to it, despite the
824 documentation. Therefore a plain malloc must be used.
826 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
827 on-stack buffers were used, resulting in a taint trap when DSN information
828 copied from a received message was written into the buffer.
830 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
831 the ordering of its ARC headers. This caused a crash.
833 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
834 a new record was being constructed with information from the peer, a trap
837 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
838 installation would get error messages from DMARC verify, when it hit the
839 nonexistent file indicated by the default. Distros wanting DMARC enabled
840 should both provide the file and set the option.
841 Also enforce no DMARC verification for command-line sourced messages.
843 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
844 could, depending on the platform, hang at the STARTTLS response.
846 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
847 message on a connection. Previously if one message had ARC headers and
848 the following one did not, a crash could result when adding an
849 Authentication-Results: header.
851 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
852 Exim main code and Exim-related utities. The introduction of taint
853 tracking also did many adjustments to string handling. Since then, eximon
854 frequently terminated with an assert failure.
856 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
857 check for 452 responses. This slightly helps the inefficieny of doing
858 a large alias-expansion into a recipient-limited target. The max_rcpt
859 transport option still applies (and at the current default, will override
860 the new feature). The check is done for either cause of synch, and forces
861 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
862 connection. The new facility is not tunable at this time.
864 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
865 library live data was being used, so the results became garbage. Make
866 copies while it is still usable.
868 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
869 on delivery deferred (==) and failed (**) lines (if a delivery was
870 attemtped). Previously it was only on completion (=>) lines.
872 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
873 for the expansion of the server_scram_iter and server_scram_salt options.
875 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
876 are now specifically given a NO_DATA response without hitting the system
877 resolver. The library goes on to do the now-standard TXT lookup.
878 Use of dnsdb lookups is not affected.
880 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
881 only retrieve the errormessage once. Previously two calls to dlerror()
882 were used, and the second one (for mainlog/paniclog) retrieved null
885 JH/20 Taint checking: disallow use of tainted data for
886 - the appendfile transport file and directory options
887 - the pipe transport command
888 - the autoreply transport file, log and once options
889 - file names used by the redirect router (including filter files)
891 - paths used by single-key lookups
892 Previously this was permitted.
894 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
895 adjusted the size of a major service buffer; this failed because the
896 buffer was in use at the time. Change to a compile-time increase in the
897 buffer size, when this authenticator is compiled into exim.
899 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
900 previous fast-mode was untenable in the face of glibs using mmap to
901 support larger malloc requests.
903 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
904 New values supported, if defined on system where compiled:
905 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
906 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
908 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
909 running a limited number of proceses in parallel, a benefit is gained. The
910 amount varies with the platform hardware and load. The use of the option
911 queue_run_in_order means we cannot do this, as ordering becomes
914 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
915 had introduced a string-copy (for ensuring NUL-termination) which was not
916 appropriate for that case, which can include embedded NUL bytes in the
917 block of data. Investigation showed the copy to actually be needless, the
918 data being length-specified.
920 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
921 done during a receiving connection, and both used TLS, global info was
922 used rather than per-connection info for tracking the state of data
923 queued for transmission. This could result in a connection hang.
925 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
926 Previously, when delivering serveral messages down a single connection
927 only the first would provide a SIZE. This was due to the size information
928 not being properly tracked.
930 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
931 TAI (at 37 seconds currently), pretend to be in UTC for time-related
932 expansion and logging. Previously, spurious values such as a future
933 minute could be seen.
935 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
936 it could crash from a null-deref. This could also affect the
937 ${addresses: } operator and ${readsock } item.
939 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
940 message following a mime one, the variable was not reset.
942 JH/30 When an pipelined-connect fails at the first response, assume incorrect
943 cached capability (perhaps the peer reneged?) and immediately retry in
946 JH/31 Fix spurious detection of timeout while writing to transport filter.
948 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
949 an attempt to copy the string was made before checking it.
951 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
952 taint of the lookup key was maintained; we now regard the presence in the
953 filesystem as sufficient validation.
955 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
956 argument is supplied.
958 JH/35 The dsearch lookup now requires that the directory is an absolute path.
959 Previously this was not checked, and nonempty relative paths made an
960 access under Exim's current working directory.
962 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
963 Previously no event was raised.
965 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
966 parameter supplied by the sender MAIL FROM command. Previously it was
967 ignored, and only the check_spool_space option value for the required
970 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
971 the size of the signing public-key. Previously it was instead giving
972 the size of the signature hash.
974 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
975 the default. See the (new) dkim_verify_min_keysizes option.
977 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
978 and an ACL use a lookup for checking either the local_part or domain,
979 stale data could be accessed. Ensure that variable references are
980 dropped between messages.
982 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
983 by the client was not checked as pointing within response data before
984 being used. A malicious client could thus cause an out-of-bounds read and
985 possibly gain authentication. Fix by adding the check.
987 JH/42 Internationalisation: change the default for downconversion in the smtp
988 transport to be "if needed". Previously it was "as previously set" for
989 the message, which usually meant "if needed" for message-submission but
990 "no" for everything else. However, MTAs have been seen using SMTPUTF8
991 even when the envelope addresses did not need it, resulting in forwarding
992 failures to non-supporting MTAs. A downconvert in such cases will be
993 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
994 transport. The change does mean that addresses needing conversion will
995 be converted when previously a delivery failure would occur.
997 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
998 response was received it would be used unchecked in a fail-DSN, violating
999 standards on line-length limits. Truncate if needed.
1001 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
1002 doesn't work. (Additionally add a new main config option to configure the
1009 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
1010 side, to decode a TLS 1.3 packet capture.
1012 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
1013 Previously the default library behaviour applied, sending two, each in
1014 its own TCP segment.
1016 JH/03 Debug output for ACL now gives the config file name and line number for
1019 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
1021 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
1023 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
1024 buffer overrun for (non-chunking) other transports.
1026 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
1027 TLS1.3, means that a server rejecting a client certificate is not visible
1028 to the client until the first read of encrypted data (typically the
1029 response to EHLO). Add detection for that case and treat it as a failed
1030 TLS connection attempt, so that the normal retry-in-clear can work (if
1031 suitably configured).
1033 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
1034 and/or domain. Found and fixed by Jason Betts.
1036 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
1037 configuration). If a CNAME target was not a wellformed name pattern, a
1040 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
1041 the OS reports them interleaved with other addresses.
1043 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
1044 used both for input and for a verify callout, both encrypted, SMTP
1045 responses being sent by the server could be lost. This resulted in
1046 dropped connections and sometimes bounces generated by a peer sending
1049 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
1050 string. Previously it was possible to cause undefined behaviour in a
1051 library routine (usually a crash). Found by "zerons".
1053 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
1056 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
1057 API was removed, so update to use the newer ones.
1059 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
1060 any timeout set, is taking a long time. Previously we would hang on to a
1061 rotated logfile "forever" if the input was arriving with long gaps
1062 (a previous attempt to fix addressed lack, for a long time, of initial
1065 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
1066 shared (NFS) environment. The length of the tempfile name is now
1067 4 + 16 ("hdr.$message_exim_id") which might break on file
1068 systems which restrict the file name length to lower values.
1069 (It was "hdr.$pid".)
1071 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1072 shared (NFS) environment.
1074 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1075 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1078 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1079 on some platforms for bit 31.
1081 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1082 to changes apparently associated with TLS1.3 handling some of the APIs
1083 previously used were either nonfunctional or inappropriate. Strings
1084 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1085 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1086 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1087 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1088 and the use of specific cipher names in the encrypted= ACL condition.
1090 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1092 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1093 verification result was not updated unless hosts_require_ocsp applied.
1095 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1096 queue_list_requires_admin set to false, non-admin users were denied the
1099 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1100 directory-of-certs mode. Previously they were advertised despite the
1103 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1104 A single TCP connection by a client will now hold a TLS connection open
1105 for multiple message deliveries, by default. Previously the default was to
1108 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1109 default. If built with the facility, DANE will be used. The facility
1110 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1112 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1113 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1114 must be defined and you must still, unless you define DISABLE_TLS, manage
1115 the the include-dir and library-file requirements that go with that
1116 choice. Non-TLS builds are still supported.
1118 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1121 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1122 default. If the platform supports and has the facility enabled, it will
1123 be requested on all coneections.
1125 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1126 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1128 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1130 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1131 success-DSN messages. Previously the From: header was always the default
1132 one for these; the option was ignored.
1134 JH/28 Fix the timeout on smtp response to apply to the whole response.
1135 Previously it was reset for every read, so a teergrubing peer sending
1136 single bytes within the time limit could extend the connection for a
1137 long time. Credit to Qualsys Security Advisory Team for the discovery.
1139 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1140 delivery address, which leaked information of the results of local
1141 forwarding. Change to the original envelope recipient address, per
1144 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1145 requested. Previously not bounce was generated and a log entry of
1146 error ignored was made.
1148 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1150 JH/32 Introduce a general tainting mechanism for values read from the input
1151 channel, and values derived from them. Refuse to expand any tainted
1152 values, to catch one form of exploit.
1154 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1155 was unused and the unexpanded text used for the test. Found and
1156 fixed by Ruben Jenster.
1158 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1159 an attempt to use a TLS library read routine dereffed a nul pointer,
1162 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1163 them in our smtp response.
1165 JH/36 Have the general router option retry_use_local_part default to true when
1166 any of the restrictive preconditions are set (to anything). Previously it
1167 was only for check_local user. The change removes one item of manual
1168 configuration which is required for proper retries when a remote router
1169 handles a subset of addresses for a domain.
1171 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1172 link count into consideration.
1174 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1175 caused the extension of big_buffer, the following lines were ignored.
1177 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1178 accordance with RFC 2308. Previously there was no expiry, so a longlived
1179 receive process (eg. due to ACL delays) versus a short SOA value could
1182 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1184 JH/39 Promote DMARC support to mainline.
1186 JH/40 Bug 2452: Add a References: header to DSNs.
1188 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1189 parameters. The relevant library call is documented as "Deprecated: This
1190 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1191 3.6.0, DH parameters are negotiated following RFC7919."
1193 HS/06 Change the default of dnssec_request_domains to "*"
1195 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1196 carried on and emitted a BDAT command, even when PIPELINING was not
1199 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1200 buffer was used for the filename, resulting in a trap when tainted
1201 arguments (eg. $domain) were used.
1203 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1204 recommended to avoid a possible server-load attack. The feature can be
1205 re-enabled via the openssl_options main cofiguration option.
1207 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1208 for version 4.90 - adding a "more data" boolean to the arguments.
1209 Bumped the ABI version number also, this having been missed previously;
1210 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1211 respect of usage of smtp_printf() by either local_scan code or libraries
1212 accessed via the ${dlfunc } expansion item. Both will need coding
1213 adjustment for any calls to smtp_printf() to match the new function
1214 signature; a FALSE value for the new argument is always safe.
1216 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1217 the file-offset (which the Linux syscall does, and exim expects); this
1218 resulted in an indefinite loop.
1220 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1221 to do ARC verification. The Authentication-Results: header line added
1222 by the configuration then had no ARC item.
1224 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1225 an incorrect mode was used for reading the result, resulting in it being
1232 JH/01 Remove code calling the customisable local_scan function, unless a new
1233 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1235 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1236 non-signal-safe functions being used.
1238 JH/03 Bug 2269: When presented with a received message having a stupidly large
1239 number of DKIM-Signature headers, disable DKIM verification to avoid
1240 a resource-consumption attack. The limit is set at twenty.
1242 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1243 report of oldest_pass in ${authres } in consequence, and separate out
1244 some descriptions of reasons for verification fail.
1246 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1247 files in the spool were present and unlocked. A queue-runner could spot
1248 them, resulting in a duplicate delivery. Fix that by doing the unlock
1249 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1250 add more error-checking on spoolfile handling while that code is being
1253 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1254 No known attacks, no CVE, this is defensive hardening.
1256 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1257 a queue-runner could start a delivery while other operations were ongoing.
1258 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1259 Found and investigated by Tim Stewart. Fix by using the open message data
1260 file handle rather than opening another, and not locally closing it (which
1261 releases a lock) for that case, while creating the temporary .eml format
1262 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1264 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1265 $sender_verify_failure/$recipient_verify_failure to "random".
1267 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1270 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1271 Previously this would segfault.
1273 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1276 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1277 like zero, since the resolver should be doing this for us, But we need one
1278 as a CNAME but no MX presence gets the CNAME returned; we need to check
1279 that doesn't point to an MX to declare it "no MX returned" rather than
1280 "error, loop". A new main option is added so the older capability of
1281 following some limited number of chain links is maintained.
1283 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1285 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1286 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1287 now-deprecated earlier definitions used only specified the range up to TLS
1288 1.2 (in the older-version library docs).
1290 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1292 JH/15 Rework TLS client-side context management. Stop using a global, and
1293 explicitly pass a context around. This enables future use of TLS for
1294 connections to service-daemons (eg. malware scanning) while a client smtp
1295 connection is using TLS; with cutthrough connections this is quite likely.
1297 JH/16 Fix ARC verification to do AS checks in reverse order.
1299 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1301 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1302 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1303 Previously the "utf8" would be re-prepended for every additional message.
1305 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1306 Previously thery were accepted, resulting in issues when attempting to
1307 forward messages to a non-supporting MTA.
1309 PP/02 Let -n work with printing macros too, not just options.
1311 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1312 one parent address was copied, and bogus data was used at delivery-logging
1313 time. Either a crash (after delivery) or bogus log data could result.
1314 Discovery and analysis by Tim Stewart.
1316 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1317 Previously if the string ended mid-character, we did not insert the
1318 promised '?' replacement.
1320 PP/04 Documentation: current string operators work on bytes, not codepoints.
1322 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1323 should pack well giving a smaller memory footprint so better caching and
1324 therefore performance. Group the declarations where this can't be done so
1325 that the byte-sized flag variables are not interspersed among pointer
1326 variables, giving a better chance of good packing by the compiler.
1328 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1329 non-null, to avoid issues with sites running BATV. Previously reports were
1330 sent with an empty envelope sender so looked like bounces.
1332 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1333 The ignore_error flag wasn't being returned from the filter subprocess so
1334 was not set for later routers. Investigation and fix by Matthias Kurz.
1336 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1337 and a msg:complete for the whole, when a message is manually removed using
1338 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1340 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1341 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1342 hope that is portable enough.
1344 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1345 requiring a known-CA anchor certificate; make it now rely entirely on the
1346 TLSA as an anchor. Checking the name on the leaf cert against the name
1347 on the A-record for the host is still done for TA (but not for EE mode).
1349 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1350 included in delivery lines for non-proxied connections, when compiled with
1351 SUPPORT_SOCKS and running with proxy logging enabled.
1353 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1354 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1355 move the existing event to fire before the normal logging of message
1356 failure so that custom logging is bracketed by normal logging.
1358 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1359 msg:fail:internal event. Developement by Matthias Kurz.
1361 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1362 far too small for todays use of crypto signatures stored there. Go all
1363 the way to the max DNS message size of 64kB, even though this might be
1364 overmuch for IOT constrained device use.
1366 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1367 copy a string over itself. The library routine is documented as not
1368 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1370 JH/32 For main options check_spool_space and check_inode_space, where the
1371 platform supports 64b integers, support more than the previous 2^31 kB
1372 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1373 the previous G, M, k.
1375 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1376 $authenticated_fail_id variable on authentication failure. Previously
1379 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1380 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1381 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1382 GNUTLS_SEC_PARAM_MEDIUM.
1384 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1385 side. Previously we would continue as if no SNI had been received.
1387 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1388 "<" character, which should be a list-separator specification, we walked
1389 off past the nul-terimation.
1391 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1392 causes) even when the retry time is not yet met. Previously they were
1393 not, meaning that when (say) an account was over-quota and temp-rejecting,
1394 and multiple senders' messages were queued, only one sender would get
1395 notified on each configured delay_warning cycle.
1397 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1399 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1400 cipher-suites, an error can be left on the stack even for a succeeding
1401 accept; this results in impossible error messages when a later operation
1404 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1405 return error codes indicating retry. Under TLS1.3 this becomes required.
1407 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1408 it only wrote the new authenticators, resulting in a lack of tracking of
1409 peer changes of ESMTP extensions until the next cache flush.
1411 JH/41 Fix the loop reading a message header line to check for integer overflow,
1412 and more-often against header_maxsize. Previously a crafted message could
1413 induce a crash of the recive process; now the message is cleanly rejected.
1415 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1416 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1422 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1423 When redis_servers is set to a list of > 1 element, and the Redis servers
1424 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1425 case of MOVED into a DEFER case instead, thus moving the query onto the
1426 next server in the list. For a cluster of N elements, all N servers must
1427 be defined in redis_servers.
1429 GF/02 Catch and remove uninitialized value warning in exiqsumm
1430 Check for existence of @ARGV before looking at $ARGV[0]
1432 JH/01 Replace the store_release() internal interface with store_newblock(),
1433 which internalises the check required to safely use the old one, plus
1434 the allocate and data copy operations duplicated in both (!) of the
1435 extant use locations.
1437 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1438 modifier. This matches the restriction on the commandline.
1440 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1441 Previously only the last row was returned.
1443 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1444 we assumed that tags in the header were well-formed, and parsed the
1445 element content after inspecting only the first char of the tag.
1446 Assumptions at that stage could crash the receive process on malformed
1449 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1450 While running the DKIM ACL we operate on the Permanent memory pool so that
1451 variables created with "set" persist to the DATA ACL. Also (at any time)
1452 DNS lookups that fail create cache records using the Permanent pool. But
1453 expansions release any allocations made on the current pool - so a dnsdb
1454 lookup expansion done in the DKIM ACL releases the memory used for the
1455 DNS negative-cache, and bad things result. Solution is to switch to the
1456 Main pool for expansions.
1457 While we're in that code, add checks on the DNS cache during store_reset,
1458 active in the testsuite.
1459 Problem spotted, and debugging aided, by Wolfgang Breyha.
1461 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1462 When none of the hosts presented to a transport match an already-open
1463 connection, close it and proceed with the list. Previously we would
1464 queue the message. Spotted by Lena with Yahoo, probably involving
1467 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1468 Previously a spurious "250 OK id=" response was appended to the proper
1471 JH/08 The "support for" informational output now, which built with Content
1472 Scanning support, has a line for the malware scanner interfaces compiled
1473 in. Interface can be individually included or not at build time.
1475 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1476 by the template makefile "src/EDITME". The "STREAM" support for an older
1477 ClamAV interface method is removed.
1479 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1480 rows affected is given instead).
1482 JH/11 The runtime Berkeley DB library version is now additionally output by
1483 "exim -d -bV". Previously only the compile-time version was shown.
1485 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1486 SMTP connection. Previously, when one had more recipients than the
1487 first, an abortive onward connection was made. Move to full support for
1488 multiple onward connections in sequence, handling cutthrough connection
1489 for all multi-message initiating connections.
1491 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1492 routers. Previously, a multi-recipient message would fail to match the
1493 onward-connection opened for the first recipient, and cause its closure.
1495 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1496 a timeout on read on a GnuTLS initiating connection, resulting in the
1497 initiating connection being dropped. This mattered most when the callout
1498 was marked defer_ok. Fix to keep the two timeout-detection methods
1501 JH/15 Relax results from ACL control request to enable cutthrough, in
1502 unsupported situations, from error to silently (except under debug)
1503 ignoring. This covers use with PRDR, frozen messages, queue-only and
1506 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1508 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1509 metadata, resulting in a crash in free().
1511 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1512 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1513 Broken also in d185889f4, with init system revamp.
1515 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1516 Previously we abruptly closed the connection after reading a malware-
1517 found indication; now we go on to read the "scan ok" response line,
1520 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1521 ACL. Previously, a crash would result.
1523 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1524 macro text after a replacement (previously it was only once per line) and
1525 by skipping builtin macros when searching for an uppercase lead character.
1527 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1528 for the build is renamed.
1530 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1531 was allocated for every new TLS startup, meaning one per message. Fix
1532 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1534 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1535 reported the original. Fix to report (as far as possible) the ACL
1536 result replacing the original.
1538 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1539 OpenSSL. Certificate information is loaded for every new TLS startup,
1540 and the resources needed to be freed.
1542 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1544 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1547 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1548 DATA response info to the (existing) per-recipient response info for
1549 the "C=" log element. It can have useful tracking info from the
1550 destination system. Patch from Simon Arlott.
1552 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1553 length value. Previously this would segfault.
1555 HS/02 Support Avast multiline protoocol, this allows passing flags to
1556 newer versions of the scanner.
1558 JH/28 Ensure that variables possibly set during message acceptance are marked
1559 dead before release of memory in the daemon loop. This stops complaints
1560 about them when the debug_store option is enabled. Discovered specifically
1561 for sender_rate_period, but applies to a whole set of variables.
1562 Do the same for the queue-runner and queue-list loops, for variables set
1563 from spool message files. Do the same for the SMTP per-message loop, for
1564 certain variables indirectly set in ACL operations.
1566 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1567 as a multi-recipient message from a mailinglist manager). The coding had
1568 an arbitrary cutoff number of characters while checking for more input;
1569 enforced by writing a NUL into the buffer. This corrupted long / fast
1570 input. The problem was exposed more widely when more pipelineing of SMTP
1571 responses was introduced, and one Exim system was feeding another.
1572 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1573 receiving system, and refused recipients seen by the sending system
1574 (propating to people being dropped from mailing lists).
1575 Discovered and pinpointed by David Carter.
1577 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1578 replaced by the ${authresults } expansion.
1580 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1582 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1583 allows proper process termination in container environments.
1585 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1586 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1588 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1589 and "err_perm", deprecated since 4.83 when the RFC-defined words
1590 "temperror" and "permerror" were introduced.
1592 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1593 transport-filters or DKIM-signing. The restriction was lost in the
1594 consolidation of verify-callout and delivery SMTP handling.
1595 Extend the restriction to also cover ARC-signing.
1597 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1598 in defer=pass mode supply a 450 to the initiator. Previously the message
1601 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1602 tls_require_ciphers is used as before.
1604 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1605 "pass_unscanned". Only tmpfails from the scanner are written to
1606 the paniclog, as they may require admin intervention (permission
1607 denied, license issues). Other scanner errors (like decompression
1608 bombs) do not cause a paniclog entry.
1610 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1611 Previously it was possible to log spurious information in receive log
1614 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1615 triggered odd behaviour from Outlook Express clients.
1617 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1620 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1621 since the IETF WG has not yet settled on that versus the original
1622 "bare" representation.
1624 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1625 Previously the millisecond value corrupted the output.
1626 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1627 corrupted the output.
1633 JH/01 Rework error string handling in TLS interface so that the caller in
1634 more cases is responsible for logging. This permits library-sourced
1635 string to be attached to addresses during delivery, and collapses
1636 pairs of long lines into single ones.
1638 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1639 during configuration. Wildcards are allowed and expanded.
1641 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1642 This permits better logging.
1644 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1645 do not close it down and have the child transport start it up again on
1646 the passed-on TCP connection. Instead, proxy the child (and any
1647 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1648 affected: the continued delivery log lines do not have any DNSSEC, TLS
1649 Certificate or OCSP information. TLS cipher information is still logged.
1651 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1652 identical IP addresses on different listening ports. Will also affect
1655 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1656 add noisy ifdef guards to special-case this sillyness.
1657 Patch from Bernd Kuhls.
1659 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1660 than 255 are no longer allowed.
1662 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1663 Disable the session-cache too, which might reduce our load. Since we
1664 currrectly use a new context for every connection, both as server and
1665 client, there is no benefit for these.
1666 GnuTLS appears to not support tickets server-side by default (we don't
1667 call gnutls_session_ticket_enable_server()) but client side is enabled
1668 by default on recent versions (3.1.3 +) unless the PFS priority string
1671 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1672 <https://reproducible-builds.org/specs/source-date-epoch/>.
1674 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1675 the check for any unsuccessful recipients did not notice the limit, and
1676 erroneously found still-pending ones.
1678 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1679 MSG_MORE. Only in-clear (not on TLS connections).
1681 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1682 transport-filter is involved we can buffer the headers in memory for
1683 creating the signature, and read the spool data file once for the
1684 signature and again for transmission.
1686 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1687 4.77 as the kernel support then wasn't solid, having issues in 64bit
1688 mode. Now, it's been long enough. Add support for FreeBSD also.
1690 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1691 case where the routing stage had gathered several addresses to send to
1692 a host before calling the transport for the first, we previously failed
1693 to close down TLS in the old transport process before passing the TCP
1694 connection to the new process. The new one sent a STARTTLS command
1695 which naturally failed, giving a failed delivery and bloating the retry
1696 database. Investigation and fix prototype from Wolfgang Breyha.
1698 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1699 false-negatives in the check that the sender had not preempted a response
1700 or prompt from Exim (running as a server), due to that code's lack of
1701 awareness of the SMTP input buffering.
1703 PP/04 Add commandline_checks_require_admin option.
1704 Exim drops privileges sanely, various checks such as -be aren't a
1705 security problem, as long as you trust local users with access to their
1706 own account. When invoked by services which pass untrusted data to
1707 Exim, this might be an issue. Set this option in main configuration
1708 AND make fixes to the calling application, such as using `--` to stop
1711 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1712 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1713 responses to those, into a single TLS record each way (this usually means
1714 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1717 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1718 affects you only if you're dancing at the edge of the param size limits.
1719 If you are, and this message makes sense to you, then: raise the
1720 configured limit or use OpenSSL 1.1. Nothing we can do for older
1723 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1724 cmdline element to get the documented default one. Previously it was
1727 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1728 are unsuable for non-message-specific reasons - eg. connection timeout,
1729 banner-time rejection.
1731 JH/16 Fix logging of delivery remote port, when specified by router, under
1734 PP/06 Repair manualroute's ability to take options in any order, even if one
1735 is the name of a transport.
1738 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1740 JH/17 Change the list-building routines interface to use the expanding-string
1741 triplet model, for better allocation and copying behaviour.
1743 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1744 Previously it was constructed the first time a possibly-matching string
1745 was met in the configuration file input during startup; now it is done
1748 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1749 compatible one, to avoid the (poorly documented) possibility of a config
1750 file in the working directory redirecting the DB files, possibly correpting
1751 some existing file. CVE-2017-10140 assigned for BDB.
1753 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1754 cache-hot. Previously, although the result was properly cached, the
1755 initial verify call returned a defer.
1757 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1758 the main verify for receipient in uncached-mode.
1760 JH/22 Retire historical build files to an "unsupported" subdir. These are
1761 defined as "ones for which we have no current evidence of testing".
1763 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1764 if present. Previously it was ignored.
1766 JH/24 Start using specified-initialisers in C structure init coding. This is
1767 a C99 feature (it's 2017, so now considered safe).
1769 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1770 if was a fixed-sized field and bitmask ops via macros; it is now more
1773 PP/07 GitHub PR 56: Apply MariaDB build fix.
1774 Patch provided by Jaroslav Škarvada.
1776 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1777 during Coverity cleanups [4.87 JH/47]
1778 Diagnosis and fix provided by Michael Fischer v. Mollard.
1780 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1781 the right size to place the terminating semicolon on its own folded
1782 line, the header hash was calculated to an incorrect value thanks to
1783 the (relaxed) space the fold became.
1785 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1786 and confused the parent.
1788 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1789 which could crash as a result. This could lead to undeliverable messages.
1791 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1794 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1795 matters on fast-turnover and PID-randomising systems, which were getting
1796 out-of-order delivery.
1798 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1799 a possibly-overlapping copy. The symptom was that "Remote host closed
1800 connection in response to HELO" was logged instead of the actual 4xx
1803 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1804 Previously only that bufferd was discarded, resulting in SYMTP command
1807 JH/32 DKIM: when a message has multiple signatures matching an identity given
1808 in dkim_verify_signers, run the dkim acl once for each. Previously only
1809 one run was done. Bug 2189.
1811 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1812 "panic the current process" to "deliberately defer". The panic log is
1813 still written with the problem list name; the mail and reject logs now
1814 get a temp-reject line for the message that was being handled, saying
1815 something like "domains check lookup or other defer". The SMTP 451
1816 message is still "Temporary local problem".
1818 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1819 A crafted sequence of BDAT commands could result in in-use memory beeing
1820 freed. CVE-2017-16943.
1822 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1823 from SMTP input. Previously it was always done; now only done for DATA
1824 and not BDAT commands. CVE-2017-16944.
1826 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1827 to the message (such as an overlong header line). Previously this was
1828 not done and we did not exit BDAT mode. Followon from the previous item
1829 though a different problem.
1835 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1836 than -2003 did; needs libidn2 in addition to libidn.
1838 JH/02 The path option on a pipe transport is now expanded before use.
1840 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1841 Patch provided by "Björn", documentation fix added too.
1843 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1844 missing a wire-to-host endian conversion.
1846 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1847 close after a BDAT command line could be taken as a following command,
1848 giving a synch failure. Fix by only checking for synch immediately
1849 before acknowledging the chunk.
1851 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1852 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1853 macro. Patches provided by Josh Soref.
1855 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1856 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1857 by IANA as a service extension. However, John Klensin suggests that we
1860 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1861 the dkim code may be unix-mode line endings rather than smtp wire-format
1862 CRLF, so prepend a CR to any bare LF.
1864 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1865 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1867 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1868 the first were themselves being wrongly included in the feed into dkim
1869 processing; with most chunk sizes in use this resulted in an incorrect
1870 body hash calculated value.
1872 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1873 DKIM signature block, for verification. Although advised against by
1874 standards it is specifically not ruled illegal.
1876 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1878 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1879 missing a body hash (the bh= tag).
1881 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1882 It seems that HAProxy sends the Proxy Protocol information in clear and
1883 only then does a TLS startup, so do the same.
1885 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1886 TCP connections (such as for Spamd) unless the daemon successfully set
1887 Fast Open mode on its listening sockets. This fixes breakage seen on
1888 too-old kernels or those not configured for Fast Open, at the cost of
1889 requiring both directions being enabled for TFO, and TFO never being used
1890 by non-daemon-related Exim processes.
1892 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1893 endings, at least on the first header line. Try to canonify any that get
1894 past that check, despite the cost.
1896 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1897 now limited to an arbitrary five deep, while parsing addresses with the
1898 strip_excess_angle_brackets option enabled.
1900 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1901 instead leave the unprompted TLS handshake in socket buffer for the
1902 TLS library to consume.
1904 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1906 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1908 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1909 region in one operation, for speed, and this leaves assigned pointers
1910 dangling. Add checks run only under the testsuite which checks all
1911 variables at a store-reset and panics on a dangling pointer; add code
1912 explicitly nulling out all the variables discovered. Fixes one known
1913 bug: a transport crash, where a dangling pointer for $sending_ip_address
1914 originally assigned in a verify callout, is re-used.
1916 PP/06 Drop '.' from @INC in various Perl scripts.
1918 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1920 PP/08 Reduce a number of compilation warnings under clang; building with
1921 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1922 should be warning-free.
1924 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1926 HS/01 Fix portability problems introduced by PP/08 for platforms where
1927 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1929 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1930 chunk. This allows us to accept broken chunked messages. We need a more
1931 general solution here.
1933 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1934 already-broken messages in the queue.
1936 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1938 JH/19 Fix reference counting bug in routing-generated-address tracking.
1944 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1945 supports it and a size is available (ie. the sending peer gave us one).
1947 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1948 years of being deprecated). The replacements are the ACLs
1949 acl_smtp_mime and acl_not_smtp_mime.
1951 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1952 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1953 or even an in-clear connection were permitted. Now, if the host lookup
1954 was dnssec and dane was requested then the host is only used if the
1955 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1956 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1957 if one fails this test.
1958 This means that a poorly-configured remote DNS will make it incommunicado;
1959 but it protects against a DNS-interception attack on it.
1961 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1962 non-noisy when a race steals the message being considered.
1964 JH/05 If main configuration option tls_certificate is unset, generate a
1965 self-signed certificate for inbound TLS connections.
1967 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1968 in rewrites and routers.
1970 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1971 and logged a warning sing 4.83; now they are a configuration file error.
1973 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1974 (lacking @domain). Apply the same qualification processing as RCPT.
1976 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1978 JH/10 Support ${sha256:} applied to a string (as well as the previous
1981 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1982 a cutthrough deliver is pending, as we always want to make a connection.
1983 This also avoids re-routing the message when later placing the cutthrough
1984 connection after a verify cache hit.
1985 Do not update it with the verify result either.
1987 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1988 when routing results in more than one destination address.
1990 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1991 signing (which inhibits the cutthrough capability). Previously only
1992 the presence of an option was tested; now an expansion evaluating as
1993 empty is permissible (obviously it should depend only on data available
1994 when the cutthrough connection is made).
1996 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1997 the relevant preceding SMTP command did not note the pipelining mode.
1999 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
2000 Previously they were not counted.
2002 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
2003 as one having no matching records. Previously we deferred the message
2004 that needed the lookup.
2006 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
2007 distinguished as "(=".
2009 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
2010 for missing MX records. Previously it only worked for missing A records.
2012 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
2014 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
2015 after the data-go-ahead and data-ack. Patch from Jason Betts.
2017 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
2018 even for a "none" policy. Patch from Tony Meyer.
2020 JH/22 Fix continued use of a connection for further deliveries. If a port was
2021 specified by a router, it must also match for the delivery to be
2024 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
2025 When one of the recipients of a message was unsuitable for the connection
2026 (has no matching addresses), we lost track of needing to mark it
2027 deferred. As a result mail would be lost.
2029 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
2031 JH/25 Decoding ACL controls is now done using a binary search; the source code
2032 takes up less space and should be simpler to maintain. Merge the ACL
2033 condition decode tables also, with similar effect.
2035 JH/26 Fix problem with one_time used on a redirect router which returned the
2036 parent address unchanged. A retry would see the parent address marked as
2037 delivered, so not attempt the (identical) child. As a result mail would
2040 JH/27 Fix a possible security hole, wherein a process operating with the Exim
2041 UID can gain a root shell. Credit to http://www.halfdog.net/ for
2042 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
2045 JH/28 Enable {spool,log} filesystem space and inode checks as default.
2046 Main config options check_{log,spool}_{inodes,space} are now
2047 100 inodes, 10MB unless set otherwise in the configuration.
2049 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
2050 Previously it only applied to the main-section connection policy
2053 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
2055 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
2056 by me. Added RFC7919 DH primes as an alternative.
2058 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
2059 are not in the system include path.
2061 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
2062 GnuTLS, when a session startup failed (eg because the client disconnected)
2063 Exim did stdio operations after fclose. This was exposed by a recent
2064 change which nulled out the file handle after the fclose.
2066 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
2067 signed directly by the cert-signing cert, rather than an intermediate
2068 OCSP-signing cert. This is the model used by LetsEncrypt.
2070 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2072 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2073 an incoming connection.
2075 HS/02 Bug 1802: Do not half-close the connection after sending a request
2078 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2079 fallback to "prime256v1".
2081 JH/34 SECURITY: Use proper copy of DATA command in error message.
2082 Could leak key material. Remotely exploitable. CVE-2016-9963.
2088 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2089 and 3.4.4 - once the server is enabled to respond to an OCSP request
2090 it does even when not requested, resulting in a stapling non-aware
2091 client dropping the TLS connection.
2093 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2094 support variable-length bit vectors. No functional change.
2096 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2097 The I= interface field on outgoing lines is now after the H= remote
2098 host field, same as incoming lines. There is a separate
2099 outgoing_interface log selector which allows you to disable the
2102 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2103 If not running log_selector +smtp_connection the mainlog would be held
2104 open indefinitely after a "too many connections" event, including to a
2105 deleted file after a log rotate. Leave the per net connection logging
2106 leaving it open for efficiency as that will be quickly detected by the
2107 check on the next write.
2109 HS/01 Bug 1671: Fix post transport crash.
2110 Processing the wait-<transport> messages could crash the delivery
2111 process if the message IDs didn't exist for some reason. When
2112 using 'split_spool_directory=yes' the construction of the spool
2113 file name failed already, exposing the same netto behaviour.
2115 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2116 mime_regex ACL conditions.
2118 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2119 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2120 to HELO, local diagnostic string.
2122 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2123 log line to debug. Even when configured with a tls authenticator many
2124 client connections are expected to not authenticate in this way, so
2125 an authenticate fail is not an error.
2127 HS/02 Add the Exim version string to the process info. This way exiwhat
2128 gives some more detail about the running daemon.
2130 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2131 matter for fast-change records such as DNSBLs.
2133 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2134 retry record, even if constant. There may be multiple transports with
2135 different interface settings and the retry behaviour needs to be kept
2138 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2140 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2142 JH/10 Bug 840: fix log_defer_output option of pipe transport
2144 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2145 hosts_randomize. This matters a lot when combined with mua_wrapper.
2147 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2148 ${quote_pgsql:<string>} operator.
2150 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2152 JH/14 Change header file name for freeradius-client. Relevant if compiling
2153 with Radius support; from the Gentoo tree and checked under Fedora.
2155 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2157 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2158 option transport_filter, meaning no filtering.
2160 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2162 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2163 defaults to "*" (all hosts). The variable is now available when not built
2164 with TLS, default unset, mainly to enable keeping the testsuite sane.
2165 If a server certificate is not supplied (via tls_certificate) an error is
2166 logged, and clients will find TLS connections fail on startup. Presumably
2167 they will retry in-clear.
2168 Packagers of Exim are strongly encouraged to create a server certificate
2169 at installation time.
2171 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2172 with the $config_file variable.
2174 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2175 in transport context, after the attempt, and per-recipient. The latter type
2176 is per host attempted. The event data is the error message, and the errno
2177 information encodes the lookup type (A vs. MX) used for the (first) host,
2178 and the trailing two digits of the smtp 4xx response.
2180 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2181 to write to mainlog (or rejectlog, paniclog) in the window between file
2182 creation and permissions/ownership being changed. Particularly affects
2183 installations where exicyclog is run as root, rather than exim user;
2184 result is that the running daemon panics and dies.
2186 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2188 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2189 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2190 "pri" and "weight". Note that the previous implicit priority given by the
2191 list order is no longer honoured.
2193 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2194 for DKIM processing.
2196 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2197 by defining SUPPORT_SOCKS.
2199 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2200 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2201 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2202 variables are renamed to proxy_{local,external}_{address,port}.
2204 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2205 for a build by defining SUPPORT_I18N
2207 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2208 of the query string, and make ${quote_redis:} do that quoting.
2210 JH/29 Move Events support from Experimental to mainline, enabled by default
2211 and removable for a build by defining DISABLE_EVENT.
2213 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2215 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2216 cached by the daemon.
2218 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2219 by defining LOOKUP_REDIS. The libhiredis library is required.
2221 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2222 keys are given for lookup.
2224 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2225 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2226 only supported when built with TLS support. The PolarSSL SHA routines
2227 are still used when the TLS library is too old for convenient support.
2229 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2230 openssl_options), for security. OpenSSL forces this from version 1.1.0
2231 server-side so match that on older versions.
2233 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2234 allocation for $value could be released as the expansion processing
2235 concluded, but leaving the global pointer active for it.
2237 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2238 and to use the domains and local_parts ACL conditions.
2240 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2241 incorrectly not doubled on cutthrough transmission, hence seen as a
2242 body-termination at the receiving system - resulting in truncated mails.
2243 Commonly the sender saw a TCP-level error, and retransmitted the message
2244 via the normal store-and-forward channel. This could result in duplicates
2245 received - but deduplicating mailstores were liable to retain only the
2246 initial truncated version.
2248 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2250 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2252 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2253 we're in there, support oversigning also; bug 1309.
2255 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2257 HS/04 Add support for keep_environment and add_environment options.
2259 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2260 either intentional arithmetic overflow during PRNG, or testing config-
2263 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2264 delivery resulted in actual delivery. Cancel cutthrough before DATA
2267 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2268 rejects a recipient: pass the reject to the originator.
2270 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2271 Many were false-positives and ignorable, but it's worth fixing the
2274 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2275 for the new environment-manipulation done at startup. Move the routines
2276 from being local to tls.c to being global via the os.c file.
2278 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2279 an extract embedded as result-arg for a map, the first arg for extract
2280 is unavailable so we cannot tell if this is a numbered or keyed
2281 extraction. Accept either.
2287 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2290 JH/02 The smtp transport option "multi_domain" is now expanded.
2292 JH/03 The smtp transport now requests PRDR by default, if the server offers
2295 JH/04 Certificate name checking on server certificates, when exim is a client,
2296 is now done by default. The transport option tls_verify_cert_hostnames
2297 can be used to disable this per-host. The build option
2298 EXPERIMENTAL_CERTNAMES is withdrawn.
2300 JH/05 The value of the tls_verify_certificates smtp transport and main options
2301 default to the word "system" to access the system default CA bundle.
2302 For GnuTLS, only version 3.0.20 or later.
2304 JH/06 Verification of the server certificate for a TLS connection is now tried
2305 (but not required) by default. The verification status is now logged by
2306 default, for both outbound TLS and client-certificate supplying inbound
2309 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2312 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2313 Status Notification (bounce) messages are now MIME format per RFC 3464.
2314 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2315 under the control of the dsn_advertise_hosts option, and routers may
2316 have a dsn_lasthop option.
2318 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2319 default, modifiable by a malware= option. The list separator for
2320 the options can now be changed in the usual way. Bug 68.
2322 JH/10 The smtp_receive_timeout main option is now expanded before use.
2324 JH/11 The incoming_interface log option now also enables logging of the
2325 local interface on delivery outgoing connections.
2327 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2328 if the interface and destination host and port all match.
2330 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2333 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2334 Patch from Andrew Lewis.
2336 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2337 now supports optional time-restrictions, weighting, and priority
2338 modifiers per server. Patch originally by <rommer@active.by>.
2340 JH/16 The spamd_address main option now supports a mixed list of local
2341 and remote servers. Remote servers can be IPv6 addresses, and
2342 specify a port-range.
2344 JH/17 Bug 68: The spamd_address main option now supports an optional
2345 timeout value per server.
2347 JH/18 Bug 1581: Router and transport options headers_add/remove can
2348 now have the list separator specified.
2350 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2353 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2356 JH/21 Support for the A6 type of dns record is withdrawn.
2358 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2359 rather than the verbs used.
2361 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2362 from 255 to 1024 chars.
2364 JH/24 Verification callouts now attempt to use TLS by default.
2366 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2367 are generic router options now. The defaults didn't change.
2369 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2370 Original patch from Alexander Shikoff, worked over by JH.
2372 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2373 files if ZCAT_COMMAND is not executable.
2375 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2377 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2379 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2380 Normally benign, it bites when the pair was led to by a CNAME;
2381 modern usage is to not canonicalize the domain to a CNAME target
2382 (and we were inconsistent anyway for A-only vs AAAA+A).
2384 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2386 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2387 when evaluating $sender_host_dnssec.
2389 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2390 $sender_helo_dnssec variable.
2392 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2394 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2396 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2398 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2399 documented as working, but never had. Support all but $spam_report.
2401 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2402 added for tls authenticator.
2404 HS/03 Add perl_taintmode main config option
2410 TL/01 When running the test suite, the README says that variables such as
2411 no_msglog_check are global and can be placed anywhere in a specific
2412 test's script, however it was observed that placement needed to be near
2413 the beginning for it to behave that way. Changed the runtest perl
2414 script to read through the entire script once to detect and set these
2415 variables, reset to the beginning of the script, and then run through
2416 the script parsing/test process like normal.
2418 TL/02 The BSD's have an arc4random API. One of the functions to induce
2419 adding randomness was arc4random_stir(), but it has been removed in
2420 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2421 function when detected.
2423 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2424 cause callback expansion.
2426 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2427 syntax errors in an expansion can be treated as a string instead of
2428 logging or causing an error, due to the internal use of bool_lax
2429 instead of bool when processing it.
2431 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2432 server certificates when making smtp deliveries.
2434 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2436 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2438 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2440 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2441 Merged patch from Sebastian Wiedenroth.
2443 JH/05 Fix results-pipe from transport process. Several recipients, combined
2444 with certificate use, exposed issues where response data items split
2445 over buffer boundaries were not parsed properly. This eventually
2446 resulted in duplicates being sent. This issue only became common enough
2447 to notice due to the introduction of connection certificate information,
2448 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2450 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2451 size buffer was used, resulting in syntax errors when an expansion
2454 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2455 version 3.3.6 or later.
2457 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2458 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2459 both become "event_action", the variables become $event_name, $event_data
2460 and $event_defer_errno. There is a new variable $verify_mode, usable in
2461 routers, transports and related events. The tls:cert event is now also
2462 raised for inbound connections, if the main configuration event_action
2465 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2466 early OCSP support, but no stapling (appears to be less than 1.0.0).
2468 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2469 server certificate names available under the smtp transport option
2470 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2473 JH/10 Time-related extraction expansions from certificates now use the main
2474 option "timezone" setting for output formatting, and are consistent
2475 between OpenSSL and GnuTLS compilations. Bug 1541.
2477 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2478 encoded parameter in the incoming message. Bug 1558.
2480 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2481 include certificate info, eximon was claiming there were spoolfile
2484 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2486 JH/14 Log delivery-related information more consistently, using the sequence
2487 "H=<name> [<ip>]" wherever possible.
2489 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2490 are problematic for Debian distribution, omit them from the release
2493 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2495 JH/16 Fix string representation of time values on 64bit time_t architectures.
2498 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2499 output list separator was used.
2504 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2505 checkers that were complaining about end of non-void function with no
2508 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2509 This was a regression introduced in 4.83 by another bugfix.
2511 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2513 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2514 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2520 TF/01 Correctly close the server side of TLS when forking for delivery.
2522 When a message was received over SMTP with TLS, Exim failed to clear up
2523 the incoming connection properly after forking off the child process to
2524 deliver the message. In some situations the subsequent outgoing
2525 delivery connection happened to have the same fd number as the incoming
2526 connection previously had. Exim would try to use TLS and fail, logging
2527 a "Bad file descriptor" error.
2529 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2530 utilities have not been installed.
2532 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2533 temporary space as the ACL may create new global variables.
2535 TL/01 LDAP support uses per connection or global context settings, depending
2536 upon the detected version of the libraries at build time.
2538 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2539 to extract and use the src ip:port in logging and expansions as if it
2540 were a direct connection from the outside internet. PPv2 support was
2541 updated based on HAProxy spec change in May 2014.
2543 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2545 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2546 Properly escape header and check for NULL return.
2548 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2551 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2553 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2554 characters in header names, implemented as a verify condition.
2555 Contributed by Michael Fischer v. Mollard.
2557 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2558 results permerror and temperror. Previous values are deprecated but
2559 still accepted. In a future release, err_perm and err_temp will be
2560 completely removed, which will be a backward incompatibility if the
2561 ACL tests for either of these two old results. Patch contributed by
2562 user bes-internal on the mailing list.
2564 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2566 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2567 selectors, in both main and reject logs.
2569 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2572 JH/07 Add malware type "sock" for talking to simple daemon.
2574 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2576 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2577 routers/transports under cutthrough routing.
2579 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2580 numbers. Touch up "bool" conditional to keep the same definition.
2582 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2584 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2586 JH/12 Expand items in router/transport headers_add or headers_remove lists
2587 individually rather than the list as a whole. Bug 1452.
2589 Required for reasonable handling of multiple headers_ options when
2590 they may be empty; requires that headers_remove items with embedded
2591 colons must have them doubled (or the list-separator changed).
2593 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2594 view the policy declared in the DMARC record. Currently, $dmarc_status
2595 is a combined value of both the record presence and the result of the
2598 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2600 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2601 dnslookup router and the smtp transport (applying to the forward
2604 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2605 of ldap servers used for a specific lookup. Patch provided by Heiko
2608 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2609 New variable $lookup_dnssec_authenticated for observability.
2611 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2612 Patch submitted by Lars Timman.
2614 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2616 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2617 Requires trusted mode and valid format message id, aborts otherwise.
2618 Patch contributed by Heiko Schlichting.
2620 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2621 certextract with support for various fields. Bug 1358.
2623 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2624 is requested by default, modifiable by smtp transport option
2627 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2628 operate on certificate variables to give certificate fingerprints
2629 Also new ${sha256:cert_variable}.
2631 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2633 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2634 Christian Aistleitner.
2636 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2638 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2639 file. Patch from Wolfgang Breyha.
2641 JH/25 Expand the coverage of the delivery $host and $host_address to
2642 client authenticators run in verify callout. Bug 1476.
2644 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2645 align with daemon_smtp_ports. Bug 72.
2647 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2648 support and error reporting did not work properly.
2650 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2651 and is readable. Patch from Andrew Colin Kissa.
2653 TL/14 Enhance documentation of ${run expansion and how it parses the
2654 commandline after expansion, particularly in the case when an
2655 unquoted variable expansion results in an empty value.
2657 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2659 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2660 by renaming away from it. Addresses GH issue 15, reported by
2663 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2665 TL/15 SECURITY: prevent double expansion in math comparison functions
2666 (can expand unsanitized data). Not remotely exploitable.
2673 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2675 PP/02 Make -n do something, by making it not do something.
2676 When combined with -bP, the name of an option is not output.
2678 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2681 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2682 $sender_host_name and config options to manage this, and basic check
2685 PP/05 DSCP support for outbound connections and control modifier for inbound.
2687 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2688 (Only plugin which currently uses this is kerberos4, which nobody should
2689 be using, but we should make it available and other future plugins might
2690 conceivably use it, even though it would break NAT; stuff *should* be
2691 using channel bindings instead).
2693 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2694 name; added for Sendmail compatibility; requires admin caller.
2695 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2696 just ignore it); requires trusted caller.
2697 Also parse but ignore: -Ac -Am -X<logfile>
2700 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2702 TL/02 Add +smtp_confirmation as a default logging option.
2704 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2705 Patch by Magnus Holmgren from 2007-02-20.
2707 TL/04 Bugzilla 1281 - Spec typo.
2708 Bugzilla 1283 - Spec typo.
2709 Bugzilla 1290 - Spec grammar fixes.
2711 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2713 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2715 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2716 mailing list by by Dmitry Isaikin.
2718 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2720 JH/02 Support "G" suffix to numbers in ${if comparisons.
2722 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2724 NM/01 Bugzilla 1197 - Spec typo
2725 Bugzilla 1196 - Spec examples corrections
2727 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2729 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2730 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2733 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2734 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2736 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2737 "acl {{name}{arg}...}", and optional args on acl condition
2740 JH/05 Permit multiple router/transport headers_add/remove lines.
2742 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2744 JH/07 Avoid using a waiting database for a single-message-only transport.
2745 Performance patch from Paul Fisher. Bugzilla 1262.
2747 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2750 JH/09 Add $headers_added variable, with content from use of ACL modifier
2751 add_header (but not yet added to the message). Bugzilla 199.
2753 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2754 Pulled from Bugzilla 817 by Wolfgang Breyha.
2756 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2758 (nb: this is the same fix as in Exim 4.80.1)
2760 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2763 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2766 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2768 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2769 advertise SMTP AUTH mechanism to us, instead of a generic
2770 protocol violation error. Also, make Exim more robust to bad
2771 data from the Dovecot auth socket.
2773 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2775 When a queue runner is handling a message, Exim first routes the
2776 recipient addresses, during which it prunes them based on the retry
2777 hints database. After that it attempts to deliver the message to
2778 any remaining recipients. It then updates the hints database using
2781 So if a recipient address works intermittently, it can get repeatedly
2782 deferred at routing time. The retry hints record remains fresh so the
2783 address never reaches the final cutoff time.
2785 This is a fairly common occurrence when a user is bumping up against
2786 their storage quota. Exim had some logic in its local delivery code
2787 to deal with this. However it did not apply to per-recipient defers
2788 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2790 This change adds a proper retry rule check during routing so that the
2791 final cutoff time is checked against the message's age. We only do
2792 this check if there is an address retry record and there is not a
2793 domain retry record; this implies that previous attempts to handle
2794 the address had the retry_use_local_parts option turned on. We use
2795 this as an approximation for the destination being like a local
2796 delivery, as in LMTP.
2798 I suspect this new check makes the old local delivery cutoff check
2799 redundant, but I have not verified this so I left the code in place.
2801 TF/02 Correct gecos expansion when From: is a prefix of the username.
2803 Test 0254 submits a message to Exim with the header
2807 When I ran the test suite under the user fanf2, Exim expanded
2808 the header to contain my full name, whereas it should have added
2809 a Resent-Sender: header. It erroneously treats any prefix of the
2810 username as equal to the username.
2812 This change corrects that bug.
2814 GF/01 DCC debug and logging tidyup
2815 Error conditions log to paniclog rather than rejectlog.
2816 Debug lines prefixed by "DCC: " to remove any ambiguity.
2818 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2820 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2821 Bug spotted by Jeremy Harris; was flawed since initial commit.
2822 Would have resulted in OCSP responses post-SNI triggering an Exim
2823 NULL dereference and crash.
2825 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2827 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2828 Bug detection, analysis and fix by Samuel Thibault.
2829 Bugzilla 1331, Debian bug #698092.
2831 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2833 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2834 Server implementation by Todd Lyons, client by JH.
2835 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2836 config variable "prdr_enable" controls whether the server
2837 advertises the facility. If the client requests PRDR a new
2838 acl_data_smtp_prdr ACL is called once for each recipient, after
2839 the body content is received and before the acl_smtp_data ACL.
2840 The client is controlled by both of: a hosts_try_prdr option
2841 on the smtp transport, and the server advertisement.
2842 Default client logging of deliveries and rejections involving
2843 PRDR are flagged with the string "PRDR".
2845 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2846 fclose(). Diagnosis by Todd Lyons.
2848 PP/17 Update configure.default to handle IPv6 localhost better.
2849 Patch by Alain Williams (plus minor tweaks).
2852 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2853 This is now consistent with GnuTLS, and is now documented: the
2854 previous undocumented portable approach to treating the option as
2855 unset was to force an expansion failure. That still works, and
2856 an empty string is now equivalent.
2858 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2859 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2860 not performing validation itself.
2862 PP/20 Added force_command boolean option to pipe transport.
2863 Patch from Nick Koston, of cPanel Inc.
2865 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2868 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2870 PP/21 Fix eximon continuous updating with timestamped log-files.
2871 Broken in a format-string cleanup in 4.80, missed when I repaired the
2872 other false fix of the same issue.
2873 Report and fix from Heiko Schlichting.
2876 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2877 Report from Prashanth Katuri.
2879 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2880 It's SecureTransport, so affects any MacOS clients which use the
2881 system-integrated TLS libraries, including email clients.
2883 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2884 using a MIME ACL for non-SMTP local injection.
2885 Report and assistance in diagnosis by Warren Baker.
2887 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2889 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2891 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2892 last id that failed so it may be referenced in subsequent ACL's.
2894 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2897 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2898 ldap library initialization, allowing self-signed CA's to be
2899 used. Also properly sets require_cert option later in code by
2900 using NULL (global ldap config) instead of ldap handle (per
2901 session). Bug diagnosis and testing by alxgomz.
2903 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2904 the src/util/ subdirectory.
2906 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2907 renamed to Transport Post Delivery Action by Jeremy Harris, as
2910 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2911 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2912 redis_servers = needs to be configured which will be used by the redis
2913 lookup. Patch from Warren Baker, of The Packet Hub.
2915 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2917 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2918 hostname or reverse DNS when processing a host list. Used suggestions
2919 from multiple comments on this bug.
2921 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2923 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2924 Missed a few lines, added it to make the runtest require no keyboard
2927 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2928 contains upper case chars. Make router use caseful_local_part.
2930 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2931 support when GnuTLS has been built with p11-kit.
2937 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2939 This, or similar/improved, will also be change PP/11 of 4.82.
2945 PP/01 Handle short writes when writing local log-files.
2946 In practice, only affects FreeBSD (8 onwards).
2947 Bugzilla 1053, with thanks to Dmitry Isaikin.
2949 NM/01 Bugzilla 949 - Documentation tweak
2951 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2954 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2956 PP/02 Implemented gsasl authenticator.
2958 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2960 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2961 `pkg-config foo` for cflags/libs.
2963 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2964 with rest of GSASL and with heimdal_gssapi.
2966 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2967 `pkg-config foo` for cflags/libs for the TLS implementation.
2969 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2970 properties get this fed in as external SSF. A number of robustness
2971 and debugging improvements to the cyrus_sasl authenticator.
2973 PP/08 cyrus_sasl server now expands the server_realm option.
2975 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2976 Patch by Jeremy Harris.
2978 PP/10 Added dbmjz lookup type.
2980 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2982 PP/12 MAIL args handles TAB as well as SP, for better interop with
2983 non-compliant senders.
2984 Analysis and variant patch by Todd Lyons.
2986 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2987 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2988 Patch from Dirk Mueller <dmueller@suse.com>
2990 PP/13 tls_peerdn now print-escaped for spool files.
2991 Observed some $tls_peerdn in wild which contained \n, which resulted
2992 in spool file corruption.
2994 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2995 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2996 or write after TLS renegotiation, which otherwise led to messages
2999 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
3000 as a tracking header (ie: a signed header comes before the signature).
3001 Patch from Wolfgang Breyha.
3003 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
3004 comma-sep list; embedded commas doubled.
3006 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
3008 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
3010 Report and patch from Dmitry Banschikov.
3012 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
3013 Removed SSL_clear() after SSL_new() which led to protocol negotiation
3014 failures. We appear to now support TLS1.1+ with Exim.
3016 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
3017 lets Exim select keys and certificates based upon TLS SNI from client.
3018 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
3019 before an outbound SMTP session. New log_selector, +tls_sni.
3021 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
3022 NULL dereference. Report and patch from Alun Jones.
3024 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
3025 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
3026 Not seeing resolver debug output on NetBSD, but suspect this is a
3027 resolver implementation change.
3029 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
3030 Left warnings. Added "eximon gdb" invocation mode.
3032 PP/21 Defaulting "accept_8bitmime" to true, not false.
3034 PP/22 Added -bw for inetd wait mode support.
3036 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
3037 locate the relevant includes and libraries. Made this the default.
3039 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
3040 Bugzilla 1246, report and most of solution from Tomasz Kusy.
3042 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
3043 This may cause build issues on older platforms.
3045 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
3046 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
3047 gnutls_require_mac & gnutls_require_protocols (no longer supported).
3048 Added SNI support via GnuTLS too.
3049 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
3051 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
3053 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
3054 Applied second patch from Janne, implementing suggestion to default
3055 multiple-strings-in-record handling to match SPF spec.
3057 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
3059 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
3060 read-only, out of scope).
3061 Patch from Wolfgang Breyha, report from Stuart Northfield.
3063 PP/29 Fix three issues highlighted by clang analyser static analysis.
3064 Only crash-plausible issue would require the Cambridge-specific
3065 iplookup router and a misconfiguration.
3066 Report from Marcin Mirosław.
3068 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3070 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3071 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3072 As part of this, removing so much warning spew let me fix some minor
3073 real issues in debug logging.
3075 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3076 assignment on my part. Fixed.
3078 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3079 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3080 Janne Snabb (who went above and beyond: thank you).
3082 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3083 string otherwise requires a connection and a bunch more work and it's
3084 relatively easy to get wrong. Should also expose TLS library linkage
3087 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3088 64-bit ${eval} (JH/03).
3090 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3091 GNU libc to support some of the 64-bit stuff, should not lead to
3092 conflicts. Defined before os.h is pulled in, so if a given platform
3093 needs to override this, it can.
3095 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3096 protection layer was required, which is not implemented.
3097 Bugzilla 1254, patch from Wolfgang Breyha.
3099 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3100 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3101 tls_dhparam take prime identifiers. Also unbreak combination of
3102 OpenSSL+DH_params+TLSSNI.
3104 PP/39 Disable SSLv2 by default in OpenSSL support.
3110 PP/01 Solaris build fix for Oracle's LDAP libraries.
3111 Bugzilla 1109, patch from Stephen Usher.
3113 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3115 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3118 TF/02 Fix a couple more cases where we did not log the error message
3119 when unlink() failed. See also change 4.74-TF/03.
3121 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3122 lock up or crash if it happened to be inside a call to libc when it
3123 got a SIGUSR1 from exiwhat.
3125 The SIGUSR1 handler appends the current process status to the process
3126 log which is later printed by exiwhat. It used to use the general
3127 purpose logging code to do this, but several functions it calls are
3128 not safe for signals.
3130 The new output code in the SIGUSR1 handler is specific to the process
3131 log, and simple enough that it's easy to inspect for signal safety.
3132 Removing some special cases also simplifies the general logging code.
3133 Removing the spurious timestamps from the process log simplifies
3136 TF/04 Improved ratelimit ACL condition.
3138 The /noupdate option has been deprecated in favour of /readonly which
3139 has clearer semantics. The /leaky, /strict, and /readonly update modes
3140 are mutually exclusive. The update mode is no longer included in the
3141 database key; it just determines when the database is updated. (This
3142 means that when you upgrade Exim will forget old rate measurements.)
3144 Exim now checks that the per_* options are used with an update mode that
3145 makes sense for the current ACL. For example, when Exim is processing a
3146 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3147 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3148 must specify per_mail/readonly. If you omit the update mode it defaults to
3149 /leaky where that makes sense (as before) or /readonly where required.
3151 The /noupdate option is now undocumented but still supported for
3152 backwards compatibility. It is equivalent to /readonly except that in
3153 ACLs where /readonly is required you may specify /leaky/noupdate or
3154 /strict/noupdate which are treated the same as /readonly.
3156 A useful new feature is the /count= option. This is a generalization
3157 of the per_byte option, so that you can measure the throughput of other
3158 aggregate values. For example, the per_byte option is now equivalent
3159 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3161 The per_rcpt option has been generalized using the /count= mechanism
3162 (though it's more complicated than the per_byte equivalence). When it is
3163 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3164 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3165 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3166 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3167 Note that using per_rcpt with a non-readonly update mode in more than
3168 one ACL will cause the recipients to be double-counted. (The per_mail
3169 and per_byte options don't have this problem.)
3171 The handling of very low rates has changed slightly. If the computed rate
3172 is less than the event's count (usually one) then this event is the first
3173 after a long gap. In this case the rate is set to the same as this event's
3174 count, so that the first message of a spam run is counted properly.
3176 The major new feature is a mechanism for counting the rate of unique
3177 events. The new per_addr option counts the number of different
3178 recipients that someone has sent messages to in the last time period. It
3179 behaves like per_rcpt if all the recipient addresses are different, but
3180 duplicate recipient addresses do not increase the measured rate. Like
3181 the /count= option this is a general mechanism, so the per_addr option
3182 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3183 example, measure the rate that a client uses different sender addresses
3184 with the options per_mail/unique=$sender_address. There are further
3185 details in the main documentation.
3187 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3189 TF/06 Removed a few PCRE remnants.
3191 TF/07 Automatically extract Exim's version number from tags in the git
3192 repository when doing development or release builds.
3194 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3195 Bugzilla 879. Patch from Paul Fisher.
3197 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3198 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3201 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3203 PP/05 Variable $av_failed, true if the AV scanner deferred.
3204 Bugzilla 1078. Patch from John Horne.
3206 PP/06 Stop make process more reliably on build failure.
3207 Bugzilla 1087. Patch from Heiko Schlittermann.
3209 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3210 Bugzilla 1089. Patch from Heiko Schlittermann.
3212 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3213 Bugzilla 1131. Patch from Holger Weiß.
3215 PP/09 Handle IPv6 addresses with SPF.
3216 Bugzilla 860. Patch from Wolfgang Breyha.
3218 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3220 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3223 PP/11 match_* no longer expand right-hand-side by default.
3224 New compile-time build option, EXPAND_LISTMATCH_RHS.
3225 New expansion conditions, "inlist", "inlisti".
3227 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3229 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3231 PP/14 fix log_write() format string regression from TF/03.
3232 Bugzilla 1152. Patch from Dmitry Isaikin.
3238 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3240 PP/02 Harmonised TLS library version reporting; only show if debugging.
3241 Layout now matches that introduced for other libraries in 4.74 PP/03.
3243 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3245 PP/04 New "dns_use_edns0" global option.
3247 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3250 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3251 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3253 TK/01 Updated PolarSSL code to 0.14.2.
3254 Bugzilla 1097. Patch from Andreas Metzler.
3256 PP/07 Catch divide-by-zero in ${eval:...}.
3257 Fixes bugzilla 1102.
3259 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3262 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3263 format-string attack -- SECURITY: remote arbitrary code execution.
3265 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3266 time unintentionally subject to list matching rules, letting the header
3267 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3268 arbitrary string expansion). This allowed for information disclosure.
3270 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3271 INT_MIN/-1 -- value coerced to INT_MAX.
3277 NM/01 Workaround for PCRE version dependency in version reporting
3280 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3281 This fixes portability to compilers other than gcc, notably
3282 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3284 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3285 makefiles for portability to HP-UX and POSIX correctness.
3287 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3288 Also via indented variable definition in the Makefile.
3289 (Debugging by Oliver Heesakkers).
3291 PP/02 Restore caching of spamd results with expanded spamd_address.
3292 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3294 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3295 Improves build reliability. Fix from: Frank Elsner
3297 NM/02 Fix wide character breakage in the rfc2047 coding
3298 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3300 NM/03 Allow underscore in dnslist lookups
3301 Fixes bug 1026. Patch from Graeme Fowler
3303 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3304 Code patches from Adam Ciarcinski of NetBSD.
3306 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3309 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3310 is logged, to avoid truncation. Patch from John Horne.
3312 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3313 Patch from Jakob Hirsch.
3315 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3316 SQL string expansion failure details.
3317 Patch from Andrey Oktyabrski.
3319 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3320 Patch from Simon Arlott.
3322 PP/09 New lookups functionality failed to compile on old gcc which rejects
3323 extern declarations in function scope.
3324 Patch from Oliver Fleischmann
3326 PP/10 Use sig_atomic_t for flags set from signal handlers.
3327 Check getgroups() return and improve debugging.
3328 Fixed developed for diagnosis in bug 927 (which turned out to be
3331 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3332 Patch from Mark Zealey.
3334 PP/12 Bugzilla 1056: Improved spamd server selection.
3335 Patch from Mark Zealey.
3337 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3338 Based on patch from Heiko Schlittermann.
3340 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3341 Patch from Uwe Doering, sign-off by Michael Haardt.
3343 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3344 variable declaration deep within a block. Bug and patch from
3347 PP/15 lookups-Makefile IRIX compatibility coercion.
3349 PP/16 Make DISABLE_DKIM build knob functional.
3351 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3352 Patch by Simon Arlott
3354 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3355 variable argument macros. Our copy now differs from upstream.
3361 TF/01 Failure to get a lock on a hints database can have serious
3362 consequences so log it to the panic log.
3364 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3365 controlled using the smtp_confirmation log selector.
3367 TF/03 Include the error message when we fail to unlink a spool file.
3369 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3370 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3371 for maintaining out-of-tree patches for some time.
3373 PP/01 Bugzilla 139: Documentation and portability issues.
3374 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3375 Handle per-OS dynamic-module compilation flags.
3377 PP/02 Let /dev/null have normal permissions.
3378 The 4.73 fixes were a little too stringent and complained about the
3379 permissions on /dev/null. Exempt it from some checks.
3380 Reported by Andreas M. Kirchwitz.
3382 PP/03 Report version information for many libraries, including
3383 Exim version information for dynamically loaded libraries. Created
3384 version.h, now support a version extension string for distributors
3385 who patch heavily. Dynamic module ABI change.
3387 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3388 privilege escalation vulnerability whereby the Exim run-time user
3389 can cause root to append content of the attacker's choosing to
3392 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3395 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3396 If dropping privileges for untrusted macros, we disabled normal logging
3397 on the basis that it would fail; for the Exim run-time user, this is not
3398 the case, and it resulted in successful deliveries going unlogged.
3399 Fixed. Reported by Andreas Metzler.
3405 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3406 only prepend for the Resent-* case. Fixes regression introduced in
3407 Exim 4.70 by NM/22 for Bugzilla 607.
3409 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3410 increasing numbers of administrators be bitten by this.
3412 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3414 PP/03 Bugzilla 994: added openssl_options main configuration option.
3416 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3418 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3420 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3422 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3423 without a peer certificate, leading to a segfault because of an
3424 assumption that peers always have certificates. Be a little more
3425 paranoid. Problem reported by Martin Tscholak.
3427 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3428 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3429 NB: ClamAV planning to remove STREAM in "middle of 2010".
3430 CL also introduces -bmalware, various -d+acl logging additions and
3431 more caution in buffer sizes.
3433 PP/09 Implemented reverse_ip expansion operator.
3435 PP/10 Bugzilla 937: provide a "debug" ACL control.
3437 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3439 PP/12 Bugzilla 973: Implement --version.
3441 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3443 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3445 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3447 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3448 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3449 ignore trailing whitespace.
3451 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3453 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3456 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3457 Notification from Dr Andrew Aitchison.
3459 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3460 ExtendedDetectionInfo response format.
3461 Notification from John Horne.
3463 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3466 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3467 XSL and documented dependency on system catalogs, with examples of how
3470 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3473 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3474 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3475 configuration file which is writeable by the Exim user or group.
3477 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3478 of configuration files to cover files specified with the -C option if
3479 they are going to be used with root privileges, not just the default
3482 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3483 option (effectively making it always true).
3485 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3486 files to be used while preserving root privileges.
3488 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3489 that rogue child processes cannot use them.
3491 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3492 run-time user, instead of root.
3494 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3495 Exim run-time user without dropping privileges.
3497 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3498 result string, instead of calling string_vformat() twice with the same
3501 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3502 for other users. Others should always drop root privileges if they use
3503 -C on the command line, even for a whitelisted configure file.
3505 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3507 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3513 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3514 $data_path, and $header_path variables; fixed documentation bugs and
3517 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3518 exipick to access non-standard spools, including the "frozen" queue
3521 NM/01 Bugzilla 965: Support mysql stored procedures.
3522 Patch from Alain Williams
3524 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3526 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3527 Patch from Andreas Metzler
3529 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3530 Patch from Kirill Miazine
3532 NM/05 Bugzilla 671: Added umask to procmail example.
3534 JJ/03 installed exipick 20100323.0, fixing doc bug
3536 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3537 directory. Notification and patch from Dan Rosenberg.
3539 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3541 TK/02 Improve log output when DKIM signing operation fails.
3543 MH/01 Treat the transport option dkim_domain as a colon separated
3544 list, not as a single string, and sign the message with each element,
3545 omitting multiple occurences of the same signer.
3547 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3548 Bugzilla 985, 986. Patch by Simon Arlott
3550 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3551 Patch by Simon Arlott
3553 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3554 MBX locking. Notification from Dan Rosenberg.
3560 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3562 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3564 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3566 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3568 NM/04 Bugzilla 915: Flag broken perl installation during build.
3574 TK/01 Added patch by Johannes Berg that expands the main option
3575 "spamd_address" if it starts with a dollar sign.
3577 TK/02 Write list of recipients to X-Envelope-Sender header when building
3578 the mbox-format spool file for content scanning (suggested by Jakob
3581 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3582 (http://www.dcc-servers.net/) support via dccifd. Activated by
3583 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3585 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3586 by Mark Daniel Reidel <mr@df.eu>.
3588 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3589 When building exim an external PCRE library is now needed -
3590 PCRE is a system library on the majority of modern systems.
3591 See entry on PCRE_LIBS in EDITME file.
3593 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3594 conversation. Added nologin parameter to request.
3595 Patch contributed by Kirill Miazine.
3597 TF/01 Do not log submission mode rewrites if they do not change the address.
3599 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3601 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3602 log files in place. Contributed by Roberto Lima.
3604 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3606 TF/03 Bugzilla 615: When checking the local_parts router precondition
3607 after a local_part_suffix or local_part_prefix option, Exim now
3608 does not use the address's named list lookup cache, since this
3609 contains cached lookups for the whole local part.
3611 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3612 Robert Millan. Documentation is in experimental-spec.txt.
3614 TF/04 Bugzilla 668: Fix parallel build (make -j).
3616 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3618 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3619 Patch provided by Jan Srzednicki.
3621 TF/05 Leading white space used to be stripped from $spam_report which
3622 wrecked the formatting. Now it is preserved.
3624 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3625 that they are available at delivery time.
3627 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3629 TF/08 TLS error reporting now respects the incoming_interface and
3630 incoming_port log selectors.
3632 TF/09 Produce a more useful error message if an SMTP transport's hosts
3633 setting expands to an empty string.
3635 NM/06 Bugzilla 744: EXPN did not work under TLS.
3636 Patch provided by Phil Pennock.
3638 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3639 Patch provided by Richard Godbee.
3641 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3642 acl_smtp_notquit, added index entry.
3644 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3645 Patch provided by Eugene Bujak.
3647 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3648 accept(). Patch provided by Maxim Dounin.
3650 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3651 Patch provided by Phil Pennock.
3653 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3655 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3656 Patch provided by Brad "anomie" Jorsch.
3658 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3659 Patch provided by Dean Brooks.
3661 TK/05 Add native DKIM support (does not depend on external libraries).
3663 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3664 Patch provided by Graeme Fowler.
3666 NM/16 Bugzilla 851: Documentation example syntax fix.
3668 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3670 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3673 NM/19 Bugzilla 745: TLS version reporting.
3674 Patch provided by Phil Pennock.
3676 NM/20 Bugzilla 167: bool: condition support.
3677 Patch provided by Phil Pennock.
3679 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3680 clients. Patch provided by Phil Pennock.
3682 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3683 Patch provided by Brad "anomie" Jorsch.
3685 NM/23 Bugzilla 687: Fix misparses in eximstats.
3686 Patch provided by Heiko Schlittermann.
3688 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3689 Patch provided by Heiko Schlittermann.
3691 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3692 plus update to original patch.
3694 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3696 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3697 Patch provided by David Brownlee.
3699 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3701 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3703 NM/30 Bugzilla 888: TLS documentation bugfixes.
3705 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3707 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3708 Unlike the original bugzilla I have changed all shell scripts in src tree.
3710 NM/33 Bugzilla 898: Transport filter timeout fix.
3711 Patch by Todd Rinaldo.
3713 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3714 Patch by Serge Demonchaux.
3716 NM/35 Bugzilla 39: Base64 decode bug fixes.
3717 Patch by Jakob Hirsch.
3719 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3721 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3723 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3725 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3731 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3732 ALT-N's libdkim that I have put here:
3733 http://duncanthrax.net/exim-experimental/
3735 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3736 were called 'true' and it seems that C99 defines that as a reserved
3737 keyword to be used with 'bool' variable types. That means you could
3738 not include C99-style headers which use bools without triggering
3739 build errors in sieve.c.
3741 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3742 as mailq or other aliases. Changed the --help handling significantly
3743 to do whats expected. exim_usage() emits usage/help information.
3745 SC/01 Added the -bylocaldomain option to eximstats.
3747 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3749 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3751 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3757 PH/01 Another patch from the Sieve maintainer.
3759 PH/02 When an IPv6 address is converted to a string for single-key lookup
3760 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3761 dots are used instead of colons so that keys in lsearch files need not
3762 contain colons. This was done some time before quoting was made available
3763 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3764 (notated using the quote facility) so as to distinguish them from IPv4
3765 keys. This meant that lookups for IP addresses in host lists did not work
3766 for iplsearch lookups.
3768 This has been fixed by arranging for IPv6 addresses to be expressed with
3769 colons if the lookup type is iplsearch. This is not incompatible, because
3770 previously such lookups could never work.
3772 The situation is now rather anomalous, since one *can* have colons in
3773 ordinary lsearch keys. However, making the change in all cases is
3774 incompatible and would probably break a number of configurations.
3776 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3779 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3780 conversion specification without a maximum field width, thereby enabling
3781 a rogue spamd server to cause a buffer overflow. While nobody in their
3782 right mind would setup Exim to query an untrusted spamd server, an
3783 attacker that gains access to a server running spamd could potentially
3784 exploit this vulnerability to run arbitrary code as the Exim user.
3786 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3787 $primary_hostname instead of what libspf2 thinks the hosts name is.
3789 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3790 a directory entry by the name of the lookup key. Previously, if a
3791 symlink pointed to a non-existing file or a file in a directory that
3792 Exim lacked permissions to read, a lookup for a key matching that
3793 symlink would fail. Now it is enough that a matching directory entry
3794 exists, symlink or not. (Bugzilla 503.)
3796 PH/03 The body_linecount and body_zerocount variables are now exported in the
3799 PH/04 Added the $dnslist_matched variable.
3801 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3802 This means they are set thereafter only if the connection becomes
3805 PH/06 Added the client_condition to authenticators so that some can be skipped
3806 by clients under certain conditions.
3808 PH/07 The error message for a badly-placed control=no_multiline_responses left
3809 "_responses" off the end of the name.
3811 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3813 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3814 (without spaces) instead of just copying the configuration text.
3816 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3818 PH/11 Added $max_received_linelength.
3820 PH/12 Added +ignore_defer and +include_defer to host lists.
3822 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3823 way in which PCRE > 7.0 is built.
3825 PH/14 Implemented queue_only_load_latch.
3827 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3828 MAIL command. The effect was to mangle the value on 64-bit systems.
3830 PH/16 Another patch from the Sieve maintainer.
3832 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3834 PH/18 If a system quota error occurred while trying to create the file for
3835 a maildir delivery, the message "Mailbox is full" was not appended to the
3836 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3837 only to a quota excession during the actual writing of the file.
3839 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3840 characters?) which causes problems in log lines. The DN values are now
3841 passed through string_printing() before being added to log lines.
3843 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3844 and InterBase are left for another time.)
3846 PH/21 Added message_body_newlines option.
3848 PH/22 Guard against possible overflow in moan_check_errorcopy().
3850 PH/23 POSIX allows open() to be a macro; guard against that.
3852 PH/24 If the recipient of an error message contained an @ in the local part
3853 (suitably quoted, of course), incorrect values were put in $domain and
3854 $local_part during the evaluation of errors_copy.
3860 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3861 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3864 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3865 issue a MAIL command.
3867 PH/02 In an ACL statement such as
3869 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3871 if a client was not listed at all, or was listed with a value other than
3872 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3873 the condition was not true (as it should be), so access was not denied.
3874 The bug was that the ! inversion was incorrectly passed on to the second
3875 item. This has been fixed.
3877 PH/03 Added additional dnslists conditions == and =& which are different from
3878 = and & when the dns lookup returns more than one IP address.
3880 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3881 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3883 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3884 FSYNC, which compiles an option called disable_fsync that allows for
3885 bypassing fsync(). The documentation is heavily laced with warnings.
3887 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3889 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3890 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3891 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3892 including adding "make clean"; (3) Added -fPIC when compiling the test
3893 dynamically loaded module, to get rid of a warning.
3895 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3896 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3897 = 0s. The bug is otherwise harmless.
3899 PH/07 There was a bug in the dovecot authenticator such that the value of
3900 $auth1 could be overwritten, and so not correctly preserved, after a
3901 successful authentication. This usually meant that the value preserved by
3902 the server_setid option was incorrect.
3904 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3906 PH/09 Installed PCRE release 7.0.
3908 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3909 run for batched SMTP input. It is now run at the start of every message
3910 in the batch. While fixing this I discovered that the process information
3911 (output by running exiwhat) was not always getting set for -bs and -bS
3912 input. This is fixed, and it now also says "batched" for BSMTP.
3914 PH/11 Added control=no_pipelining.
3916 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3917 patch, slightly modified), and move the expansion of helo_data till after
3918 the connection is made in the smtp transport (so it can use these
3921 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3923 PH/14 Added log_selector = +pid.
3925 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3927 PH/16 Add ${if forany and ${if forall.
3929 PH/17 Added dsn_from option to vary the From: line in DSNs.
3931 PH/18 Flush SMTP output before performing a callout, unless control =
3932 no_callout_flush is set.
3934 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3935 was true (the default) a successful delivery failed to delete the retry
3936 item, thus causing premature timeout of the address. The bug is now
3939 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3941 PH/21 Long custom messages for fakedefer and fakereject are now split up
3942 into multiline responses in the same way that messages for "deny" and
3943 other ACL rejections are.
3945 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3946 with slight modification.
3948 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3949 draft, changing the syntax and factoring some duplicate code".
3951 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3952 for deliveries of the second and subsequent messages over the same SMTP
3955 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3956 ${reduce, with only minor "tidies".
3958 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3960 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3961 expansion side effects.
3963 PH/27 When a message times out after an over-quota error from an Exim-imposed
3964 quota, the bounce message says "mailbox is full". This message was not
3965 being given when it was a system quota that was exceeded. It now should
3968 MH/03 Made $recipients available in local_scan(). local_scan() already has
3969 better access to the recipient list through recipients_list[], but
3970 $recipients can be useful in postmaster-provided expansion strings.
3972 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3973 in the case of a MAIL command with additional options following the
3974 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3975 were accidentally chopped off.
3977 PH/29 SMTP synchronization checks are implemented when a command is read -
3978 there is a check that no more input is waiting when there shouldn't be
3979 any. However, for some commands, a delay in an ACL can mean that it is
3980 some time before the response is written. In this time, more input might
3981 arrive, invalidly. So now there are extra checks after an ACL has run for
3982 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3983 pipelining has not been advertised.
3985 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3987 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3988 returned in the SMTP response when smtp_return_error_details was false.
3989 This has been fixed.
3991 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3992 instead of the C I/O that was originally supplied, because problems were
3993 reported on Solaris.
3995 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3996 Exim which did not show up earlier: it was assuming that a call to
3997 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3998 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3999 was a macro that became an assignment, so it seemed to work. This has
4000 changed to a proper function call with a void return, hence the compile
4001 error. Exim's code has been fixed.
4003 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
4006 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
4008 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
4010 PH/37 If a message is not accepted after it has had an id assigned (e.g.
4011 because it turns out to be too big or there is a timeout) there is no
4012 "Completed" line in the log. When some messages of this type were
4013 selected by exigrep, they were listed as "not completed". Others were
4014 picked up by some special patterns. I have improved the selection
4015 criteria to be more general.
4017 PH/38 The host_find_failed option in the manualroute router can now be set
4018 to "ignore", to completely ignore a host whose IP address cannot be
4019 found. If all hosts are ignored, the behaviour is controlled by the new
4020 host_all_ignored option.
4022 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
4023 homing or because of multiple MX records with /mx) generated more than
4024 one IP address, and the following item turned out to be the local host,
4025 all the secondary addresses of the first item were incorrectly removed
4026 from the list, along with the local host and any following hosts (which
4027 is what is supposed to happen).
4029 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
4030 whoever called Exim into the -H file. In the case of the daemon it was
4031 behaving confusingly. When first started, it used values for whoever
4032 started the daemon, but after a SIGHUP it used the Exim user (because it
4033 calls itself on a restart). I have changed the code so that it now always
4036 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
4037 message are rejected with the same error (e.g. no authentication or bad
4038 sender address), and a DATA command is nevertheless sent (as can happen
4039 with PIPELINING or a stupid MUA), the error message that was given to the
4040 RCPT commands is included in the rejection of the DATA command. This is
4041 intended to be helpful for MUAs that show only the final error to their
4044 PH/42 Another patch from the Sieve maintainer.
4046 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
4047 Eximstats - Fixed some broken HTML links and added missing column headers
4049 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
4050 columns for Rejects, Temp Rejects, Ham, and Spam rows.
4052 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
4054 PH/43 Yet another patch from the Sieve maintainer.
4056 PH/44 I found a way to check for a TCP/IP connection going away before sending
4057 the response to the final '.' that terminates a message, but only in the
4058 case where the client has not sent further data following the '.'
4059 (unfortunately, this is allowed). However, in many cases there won't be
4060 any further data because there won't be any more messages to send. A call
4061 to select() can be used: if it shows that the input is "ready", there is
4062 either input waiting, or the socket has been closed. An attempt to read
4063 the next input character can distinguish the two cases. Previously, Exim
4064 would have sent an OK response which the client would never have see.
4065 This could lead to message repetition. This fix should cure that, at
4066 least in a lot of common cases.
4068 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4069 advertised in response to EHLO.
4075 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4076 fixed by 4.65/MH/01 (is this a record?) are fixed:
4078 (i) An empty string was always treated as zero by the numeric comparison
4079 operators. This behaviour has been restored.
4081 (ii) It is documented that the numeric comparison operators always treat
4082 their arguments as decimal numbers. This was broken in that numbers
4083 starting with 0 were being interpreted as octal.
4085 While fixing these problems I realized that there was another issue that
4086 hadn't been noticed. Values of message_size_limit (both the global option
4087 and the transport option) were treated as octal if they started with 0.
4088 The documentation was vague. These values are now always treated as
4089 decimal, and I will make that clear in the documentation.
4095 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4096 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4099 MH/01 Don't check that the operands of numeric comparison operators are
4100 integers when their expansion is in "skipping" mode (fixes bug
4101 introduced by 4.64-PH/07).
4103 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4104 child addresses, Exim now panics and dies. Previously, because the count
4105 is held in a short int, deliveries were likely to be lost. As such a
4106 large number of recipients for a single message is ridiculous
4107 (performance will be very, very poor), I have chosen to impose a limit
4108 rather than extend the field.
4114 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4115 leftover -K file (the existence of which was triggered by #402).
4116 While we were at it, introduced process PID as part of the -K
4117 filename. This should rule out race conditions when creating
4120 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4121 processing considerably. Previous code took too long for large mails,
4122 triggering a timeout which in turn triggers #401.
4124 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4125 in the DK code in transports.c. sendfile() is not really portable,
4126 hence the _LINUX specificness.
4128 TF/01 In the add_headers option to the mail command in an Exim filter,
4129 there was a bug that Exim would claim a syntax error in any
4130 header after the first one which had an odd number of characters
4133 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4134 callout verification, Exim cached a "reject" for the entire domain. This
4135 is correct for most verifications, but it is not correct for a recipient
4136 verification with use_sender or use_postmaster set, because in that case
4137 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4138 case of MAIL FROM:<> rejection from other early rejections (e.g.
4139 rejection of HELO). When verifying a recipient using a non-null MAIL
4140 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4141 Whatever the result of the callout, the value of the domain cache is
4142 left unchanged (for any other kind of callout, getting as far as trying
4143 RCPT means that the domain itself is ok).
4145 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4148 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4149 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4150 session, the daemon ignores ECONNECT errors and logs others; it now
4151 ignores EPIPE as well.
4153 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4154 (quoted-printable decoding).
4156 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4157 later the small subsequent patch to fix an introduced bug.
4159 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4161 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4163 PH/08 An error is now given if message_size_limit is specified negative.
4165 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4166 to be given (somewhat) arbitrary names.
4168 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4171 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4172 miscellaneous code fixes
4174 PH/10 Added the log_reject_target ACL modifier to specify where to log
4177 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4178 hostname. This is wrong, because it relates to the incoming message (and
4179 probably the interface on which it is arriving) and not to the outgoing
4180 callout (which could be using a different interface). This has been
4181 changed to use the value of the helo_data option from the smtp transport
4182 instead - this is what is used when a message is actually being sent. If
4183 there is no remote transport (possible with a router that sets up host
4184 addresses), $smtp_active_hostname is used.
4186 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4187 tweaks were necessary in order to get it to work (see also 21 below):
4188 (a) The code assumed that strncpy() returns a negative number on buffer
4189 overflow, which isn't the case. Replaced with Exim's string_format()
4191 (b) There were several signed/unsigned issues. I just did the minimum
4192 hacking in of casts. There is scope for a larger refactoring.
4193 (c) The code used strcasecmp() which is not a standard C function.
4194 Replaced with Exim's strcmpic() function.
4195 (d) The code set only $1; it now sets $auth1 as well.
4196 (e) A simple test gave the error "authentication client didn't specify
4197 service in request". It would seem that Dovecot has changed its
4198 interface. Fortunately there's a specification; I followed it and
4199 changed what the client sends and it appears to be working now.
4201 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4204 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4205 address A is aliased to B and C, where B exists and C does not. Without
4206 -v the output is "A verified" because verification stops after a
4207 successful redirection if more than one address is generated. However,
4208 with -v the child addresses are also verified. Exim was outputting "A
4209 failed to verify" and then showing the successful verification for C,
4210 with its parentage. It now outputs "B failed to verify", showing B's
4211 parentage before showing the successful verification of C.
4213 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4214 look up a TXT record in a specific list after matching in a combined
4217 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4218 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4219 they consult the DNS. I had assumed they would set it the way they
4220 wanted; and indeed my experiments on Linux seem to show that in some
4221 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4222 To be on the safe side, however, I have now made the interface to
4223 host_find_byname() similar to host_find_bydns(), with an argument
4224 containing the DNS resolver options. The host_find_byname() function now
4225 sets these options at its start, just as host_find_bydns() does. The smtp
4226 transport options dns_qualify_single and dns_search_parents are passed to
4227 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4228 of host_find_byname() use the default settings of RES_DEFNAMES
4229 (qualify_single) but not RES_DNSRCH (search_parents).
4231 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4232 spool_read_header() do less string testing, by means of a preliminary
4233 switch on the second character of optional "-foo" lines. (This is
4234 overdue, caused by the large number of possibilities that now exist.
4235 Originally there were few.) While I was there, I also converted the
4236 str(n)cmp tests so they don't re-test the leading "-" and the first
4237 character, in the hope this might squeeze out yet more improvement.
4239 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4240 flag allowing group syntax was set by the header_syntax check but not
4241 turned off, possible causing trouble later; (2) The flag was not being
4242 set at all for the header_verify test, causing "group"-style headers to
4243 be rejected. I have now set it in this case, and also caused header_
4244 verify to ignore an empty address taken from a group. While doing this, I
4245 came across some other cases where the code for allowing group syntax
4246 while scanning a header line wasn't quite right (mostly, not resetting
4247 the flag correctly in the right place). These bugs could have caused
4248 trouble for malformed header lines. I hope it is now all correct.
4250 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4251 with the "reply" argument non-NULL. The code, however (which originally
4252 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4253 but it didn't always do it. This confused somebody who was copying the
4254 code for some other use. I have removed all the tests.
4256 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4257 feature that was used to support insecure browsers during the U.S. crypto
4258 embargo. It requires special client support, and Exim is probably the
4259 only MTA that supported it -- and would never use it because real RSA is
4260 always available. This code has been removed, because it had the bad
4261 effect of slowing Exim down by computing (never used) parameters for the
4262 RSA_EXPORT functionality.
4264 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4265 authenticator to fail if there's a tab character in the incoming data
4266 (there should never be unless someone is messing about, as it's supposed
4267 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4268 if the connection is using TLS or if the remote IP is the same as the
4269 local IP, and the "valid-client-cert option" if a client certificate has
4272 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4273 authenticators. This can be used for authorization after authentication
4274 succeeds. (In the case of plaintext, it servers for both authentication
4277 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4278 if any retry times were supplied.
4280 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4281 connection, where there is no client IP address to check. In this
4282 situation, the verify now always succeeds.
4284 PH/25 Applied John Jetmore's -Mset patch.
4286 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4288 PH/27 In a string expansion for a processed (not raw) header when multiple
4289 headers of the same name were present, leading whitespace was being
4290 removed from all of them, but trailing whitespace was being removed only
4291 from the last one. Now trailing whitespace is removed from each header
4292 before concatenation. Completely empty headers in a concatenation (as
4293 before) are ignored.
4295 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4296 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4298 PH/29 [Removed. This was a change that I later backed out, and forgot to
4299 correct the ChangeLog entry (that I had efficiently created) before
4300 committing the later change.]
4302 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4303 address errors (4xx response to RCPT) over the same connection as other
4304 messages routed to the same hosts. Such deliveries are always "forced",
4305 so retry times are not inspected. This resulted in far too many retries
4306 for the affected addresses. The effect occurred only when there were more
4307 hosts than the hosts_max_try setting in the smtp transport when it had
4308 the 4xx errors. Those hosts that it had tried were not added to the list
4309 of hosts for which the message was waiting, so if all were tried, there
4310 was no problem. Two fixes have been applied:
4312 (i) If there are any address or message errors in an SMTP delivery, none
4313 of the hosts (tried or untried) are now added to the list of hosts
4314 for which the message is waiting, so the message should not be a
4315 candidate for sending over the same connection that was used for a
4316 successful delivery of some other message. This seems entirely
4317 reasonable: after all the message is NOT "waiting for some host".
4318 This is so "obvious" that I'm not sure why it wasn't done
4319 previously. Hope I haven't missed anything, but it can't do any
4320 harm, as the worst effect is to miss an optimization.
4322 (ii) If, despite (i), such a delivery is accidentally attempted, the
4323 routing retry time is respected, so at least it doesn't keep
4324 hammering the server.
4326 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4327 in ${readsocket because some servers need this prod.
4329 PH/32 Added some extra debug output when updating a wait-xxx database.
4331 PH/33 The hint "could be header name not terminated by colon", which has been
4332 given for certain expansion errors for a long time, was not being given
4333 for the ${if def:h_colon_omitted{... case.
4335 PH/34 The spec says: "With one important exception, whenever a domain list is
4336 being scanned, $domain contains the subject domain." There was at least
4337 one case where this was not true.
4339 PH/35 The error "getsockname() failed: connection reset by peer" was being
4340 written to the panic log as well as the main log, but it isn't really
4341 panic-worthy as it just means the connection died rather early on. I have
4342 removed the panic log writing for the ECONNRESET error when getsockname()
4345 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4346 runs only) independently of the message's sender address. This meant
4347 that, if the 4xx error was in fact related to the sender, a different
4348 message to the same recipient with a different sender could confuse
4349 things. In particular, this can happen when sending to a greylisting
4350 server, but other circumstances could also provoke similar problems.
4351 I have changed the default so that the retry time for these errors is now
4352 based a combination of the sender and recipient addresses. This change
4353 can be overridden by setting address_retry_include_sender=false in the
4356 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4357 remote server are returned as part of bounce messages. This was not
4358 happening for LMTP over a pipe (the lmtp transport), but now it is the
4359 same for both kinds of LMTP.
4361 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4362 in header lines that were in fact CNAMEs. This is no longer the case.
4364 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4365 and queue runs started by the daemon processed all messages. This has
4366 been fixed so that -R and -S can now usefully be given with -q<time>.
4368 PH/40 Import PCRE release 6.7 (fixes some bugs).
4370 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4372 PH/42 Give an error if -q is specified more than once.
4374 PH/43 Renamed the variables $interface_address and $interface_port as
4375 $received_ip_address and $received_port, to make it clear that these
4376 values apply to message reception, and not to the outgoing interface when
4377 a message is delivered. (The old names remain recognized, of course.)
4379 PH/44 There was no timeout on the connect() call when using a Unix domain
4380 socket in the ${readsocket expansion. There now is.
4382 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4383 be meaningful with "accept".
4385 SC/01 Eximstats V1.43
4386 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4388 SC/02 Eximstats V1.44
4389 Use a glob alias rather than an array ref in the generated
4390 parser. This improves both readability and performance.
4392 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4393 Collect SpamAssassin and rejection statistics.
4394 Don't display local sender or destination tables unless
4395 there is data to show.
4396 Added average volumes into the top table text output.
4398 SC/04 Eximstats V1.46
4399 Collect data on the number of addresses (recipients)
4400 as well as the number of messages.
4402 SC/05 Eximstats V1.47
4403 Added 'Message too big' to the list of mail rejection
4404 reasons (thanks to Marco Gaiarin).
4406 SC/06 Eximstats V1.48
4407 Mainlog lines which have GMT offsets and are too short to
4408 have a flag are now skipped.
4410 SC/07 Eximstats V1.49 (Alain Williams)
4411 Added the -emptyok flag.
4413 SC/08 Eximstats V1.50
4414 Fixes for obtaining the IP address from reject messages.
4416 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4417 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4418 whitespace changes from 4.64-PH/27
4420 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4423 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4424 are found, allow negative numbers in numeric criteria)
4426 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4428 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4429 to match changes made in 4.64-PH/43
4431 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4433 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4434 "domain" is actually an IP address, and force a failure. This locks out
4435 those revolvers/nameservers that support "A-for-A" lookups, in
4436 contravention of the specifications.
4438 PH/48 When a host name was looked up from an IP address, and the subsequent
4439 forward lookup of the name timed out, the host name was left in
4440 $sender_host_name, contrary to the specification.
4442 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4443 restricted to single-key lookups, Exim was not diagnosing an error if
4444 * or *@ was used with a query-style lookup.
4446 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4448 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4449 long ago, but noone interested enough thought of it. Let's just say that
4450 the "1.1" means that there are some new functions that weren't there at
4451 some point in the past.
4453 PH/51 Error processing for expansion failure of helo_data from an smtp
4454 transport during callout processing was broken.
4456 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4457 tested/used via the -bh/-bhc/-bs options.
4459 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4460 bug, fixed in subsequent PCRE releases).
4462 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4463 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4465 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4471 SC/01 Use a glob alias rather than an array ref in eximstats generated
4472 parser. This improves both readability and performance.
4474 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4475 Don't display local sender or destination tables in eximstats unless
4476 there is data to show.
4477 Added average volumes into the eximstats top table text output.
4479 SC/03 Collect data on the number of addresses (recipients) as well
4480 as the number of messages in eximstats.
4482 TF/01 Correct an error in the documentation for the redirect router. Exim
4483 does (usually) call initgroups() when daemonizing.
4485 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4486 with consistent privilege compared to when running as a daemon.
4488 TF/03 Note in the spec that $authenticated_id is not set for local
4489 submissions from trusted users.
4491 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4492 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4494 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4495 by adding some example configuration directives to the default
4496 configuration file. A little bit of work is required to uncomment the
4497 directives and define how usernames and passwords are checked, but
4498 there is now a framework to start from.
4500 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4501 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4502 without this. I don't know how relevant this is to other LDAP libraries.
4504 PH/02 Add the verb name to the "unknown ACL verb" error.
4506 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4508 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4510 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4511 directory not expanded when it should be if an expanded home directory
4512 was set for the address (which is overridden by the transport).
4514 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4517 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4518 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4519 because it is too late at that time, and has no effect.
4521 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4522 security issue with \' (bugzilla #107). I could not use the
4523 PQescapeStringConn() function, because it needs a PGconn value as one of
4526 PH/08 When testing addresses using -bt, indicate those final addresses that
4527 are duplicates that would not cause an additional delivery. At least one
4528 person was confused, thinking that -bt output corresponded to deliveries.
4529 (Suppressing duplicates isn't a good idea as you lose the information
4530 about possibly different redirections that led to the duplicates.)
4532 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4533 systems where poll() doesn't work, in particular OS X.
4535 PH/10 Added more information to debugging output for retry time not reached.
4537 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4538 operations in malware.c.
4540 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4543 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4544 syslog_duplication set false, log lines that would normally be written
4545 both the the main log and to the reject log were not written to syslog at
4548 PH/14 In the default configuration, change the use of "message" in ACL warn
4549 statements to "add_header".
4551 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4552 not followed by a command (e.g. "seen endif").
4554 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4555 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4558 PH/17 Added extra conditions to the default value of delay_warning_condition
4562 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4563 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4564 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4567 The Auto-Submitted: and various List- headers are standardised, whereas I
4568 don't think Precedence: ever was.
4570 PH/18 Refactored debugging code in route_finduser() to show more information,
4571 in particular, the error code if getpwnam() issues one.
4573 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4574 This is apparently needed in addition to the PH/07 change above to avoid
4575 any possible encoding problems.
4577 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4578 but not after initializing Perl.
4580 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4581 output them only if debugging. By default they are written stderr,
4582 apparently, which is not desirable.
4584 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4587 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4590 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4592 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4593 authenticated or an ident call has been made. Suppress the default
4594 values for $authenticated_id and $authenticated_sender (but permit -oMai
4595 and -oMas) when testing with -bh.
4597 PH/24 Re-jigged the order of the tests in the default configuration so that the
4598 tests for valid domains and recipients precede the DNS black list and CSA
4599 tests, on the grounds that those ones are more expensive.
4601 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4602 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4603 This bug exists in every version of Exim that I still have, right back to
4606 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4607 However, an attempt to turn on case-sensitivity in a regex key by
4608 including (?-i) didn't work because the subject string was already
4609 lowercased, and the effects were non-intuitive. It turns out that a
4610 one-line patch can be used to allow (?-i) to work as expected.
4616 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4617 other effects) broke the use of negated acl sub-conditions.
4619 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4622 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4623 "Deny" causes Exim to reject the incoming connection with a 554 error.
4624 Unfortunately, if there is a major crisis, such as a disk failure,
4625 tcp-wrappers gives "deny", whereas what one would like would be some
4626 kind of temporary error. A kludge has been added to help with this.
4627 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4628 554 error is used if errno is still zero or contains ENOENT (which occurs
4629 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4632 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4634 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4635 errors. Otherwise a message that provokes a temporary error (when other
4636 messages do not) can cause a whole host to time out.
4638 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4639 addresses were routed directly to files or pipes from a redirect router.
4640 File deliveries just didn't batch; pipe deliveries might have suffered
4643 PH/06 A failure to get a lock for a hints database would erroneously always say
4644 "Failed to get write lock", even when it was really a read lock.
4646 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4647 of 0600. This has been changed to use the value of the lockfile_mode
4648 option (which defaults to 0600).
4650 PH/08 Applied small patch from the Sieve maintainer.
4652 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4653 folder from quota calculations, a direct delivery into this folder messed
4654 up the contents of the maildirsize file. This was because the regex was
4655 used only to exclude .Trash (or whatever) when the size of the mailbox
4656 was calculated. There was no check that a delivery was happening into an
4657 excluded directory. This bug has been fixed by ignoring all quota
4658 processing for deliveries into excluded directories.
4660 PH/10 Added the maildirfolder_create_regex option to appendfile.
4666 PH/01 The code for finding all the local interface addresses on a FreeBSD
4667 system running IPv6 was broken. This may well have applied to all BSD
4668 systems, as well as to others that have similar system calls. The broken
4669 code found IPv4 interfaces correctly, but gave incorrect values for the
4670 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4671 that it would not match correctly against @[] and not recognize the IPv6
4674 PH/02 The ipliteral router was not recognizing addresses of the form user@
4675 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4677 PH/03 Added disable_ipv6.
4679 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4680 decoded form, because it is most often used to construct To: headers
4681 lines in autoreplies, and the decoded form may well be syntactically
4682 invalid. However, $reply_address has leading white space removed, and all
4683 newlines turned into spaces so that the autoreply transport does not
4686 PH/05 If group was specified without a user on a router, and no group or user
4687 was specified on a transport, the group from the router was ignored.
4689 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4690 for visible compile-time settings that can be used to change these
4691 numbers, for those that want even more. Backwards compatibility with old
4692 spool files has been maintained. However, going back to a previous Exim
4693 release will lost any variables that are in spool files.
4695 PH/07 Two small changes when running in the test harness: increase delay when
4696 passing a TCP/IP connection to a new process, in case the original
4697 process has to generate a bounce, and remove special handling of
4698 127.0.0.2 (sic), which is no longer necessary.
4700 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4701 be the same on different OS.
4703 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4706 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4707 whether --show-vars was specified or not
4709 JJ/02 exipick: Added support for new ACL variable spool format introduced
4712 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4713 syntactically invalid From: or Reply-to: line, and a filter used this to
4714 generate an autoreply, and therefore failed to obtain an address for the
4715 autoreply, Exim could try to deliver to a non-existent relative file
4716 name, causing unrelated and misleading errors. What now happens is that
4717 it logs this as a hard delivery error, but does not attempt to create a
4720 PH/11 The exinext utility has a -C option for testing purposes, but although
4721 the given file was scanned by exinext itself; it wasn't being passed on
4722 when Exim was called.
4724 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4725 an end-of-file indication when reading a command response.
4727 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4728 compiled. In many other places in Exim, IPv6 addresses are always
4729 recognized, so I have changed this. It also means that IPv4 domain
4730 literals of the form [IPV4:n.n.n.n] are now always recognized.
4732 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4733 used if the router is not running as root, for example, when verifying at
4734 ACL time, or when using -bh. The debugging output from this situation was
4735 non-existent - all you got was a failure to exec. I have made two
4738 (a) Failures to set uid/gid, the current directory, or a process leader
4739 in a subprocess such as that created by queryprogram now generate
4740 suitable debugging output when -d is set.
4742 (b) The queryprogram router detects when it is not running as root,
4743 outputs suitable debugging information if -d is set, and then runs
4744 the subprocess without attempting to change uid/gid.
4746 PH/15 Minor change to Makefile for building test_host (undocumented testing
4749 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4750 additional section of a DNS packet that returns MX or SRV records.
4751 Instead, it always explicitly searches for A/AAAA records. This avoids
4752 major problems that occur when a DNS server includes only records of one
4753 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4754 fixed another bug: if SRV records were looked up and the corresponding
4755 address records were *not* found in the additional section, the port
4756 values from the SRV records were lost.
4758 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4759 using the correct key (the original address) when searching the retry
4760 rules in order to find which one to use for generating the retry hint.
4762 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4763 adding the default one. Similarly, if it contains a Reply-To: header, the
4764 errors_reply_to option, if set, is not used.
4766 PH/19 When calculating a retry time, Exim used to measure the "time since
4767 failure" by looking at the "first failed" field in the retry record. Now
4768 it does not use this if it is later than than the arrival time of the
4769 message. Instead it uses the arrival time. This makes for better
4770 behaviour in cases where some deliveries succeed, thus re-setting the
4771 "first failed" field. An example is a quota failure for a huge message
4772 when small messages continue to be delivered. Without this change, the
4773 "time since failure" will always be short, possible causing more frequent
4774 delivery attempts for the huge message than are intended.
4775 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4777 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4778 $1, $2, $3) because the numerical variables can be reset during some
4779 expansion items (e.g. "match"), thereby losing the authentication data.
4781 PH/21 Make -bV show the size of off_t variables so that the test suite can
4782 decide whether to run tests for quotas > 2G.
4784 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4785 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4786 filecount value is greater than 2G or if a quota value is greater than 2G
4787 on a system where the size of off_t is not greater than 4, a panic error
4790 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4791 never match. The debug and -bh output now contains an explicit error
4792 message indicating a malformed IPv4 address or mask.
4794 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4795 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4796 PH/23 above applies.
4798 PH/25 Do not write to syslog when running in the test harness. The only
4799 occasion when this arises is a failure to open the main or panic logs
4800 (for which there is an explicit test).
4802 PH/26 Added the /no_tell option to "control=freeze".
4804 PH/27 If a host name lookup failed very early in a connection, for example, if
4805 the IP address matched host_lookup and the reverse lookup yielded a name
4806 that did not have a forward lookup, an error message of the form "no IP
4807 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4808 could be logged. Now it outputs the IP address instead of "NULL".
4810 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4811 allows the sender and the authenticated sender to be set when
4812 submitting a message from within Exim. Since child_open_exim() is
4813 documented for local_scan(), the new function should be too.
4815 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4816 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4817 results in an empty string is now treated as unset.
4819 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4821 PH/31 Added sender_verify_fail logging option.
4823 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4824 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4825 tidied the source and removed it altogether.
4827 PH/33 When a queue run was abandoned because the load average was too high, a
4828 log line was always written; now it is written only if the queue_run log
4829 selector is set. In addition, the log line for abandonment now contains
4830 information about the queue run such as the pid. This is always present
4831 in "start" and "stop" lines but was omitted from the "abandon" line.
4833 PH/34 Omit spaces between a header name and the colon in the error message that
4834 is given when verify = headers_syntax fails (if there are lots of them,
4835 the message gets confusing).
4837 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4838 names, as there are now some PTR records that contain slashes. This check
4839 is only to protect against broken name servers that fall over on strange
4840 characters, so the fact that it applies to all lookups doesn't matter.
4842 PH/36 Now that the new test suite is complete, we can remove some of the
4843 special code in Exim that was needed for the old test suite. For example,
4844 sorting DNS records because real resolvers return them in an arbitrary
4845 order. The new test suite's fake resolver always returns records in the
4848 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4849 bounces) except when queue_only is set, to avoid logging races between
4850 the different processes.
4852 PH/38 Panic-die if .include specifies a non-absolute path.
4854 PH/39 A tweak to the "H" retry rule from its user.
4856 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4857 a label. They prevented compilation on older perls.
4859 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4860 a warning to be raised on newish perls.
4862 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4863 on queue. Changes to match documented behaviour of showing count of
4864 messages matching specified criteria.
4866 PH/40 Changed the default ident timeout from 30s to 5s.
4868 PH/41 Added support for the use of login_cap features, on those BSD systems
4869 that have them, for controlling the resources used by pipe deliveries.
4871 PH/42 The content-scanning code uses fopen() to create files in which to put
4872 message data. Previously it was not paying any attention to the mode of
4873 the files. Exim runs with umask(0) because the rest of the code creates
4874 files with open(), and sets the required mode explicitly. Thus, these
4875 files were ending up world-writeable. This was not a big issue, because,
4876 being within the spool directory, they were not world-accessible. I have
4877 created a function called modefopen, which takes an additional mode
4878 argument. It sets umask(777), creates the file, chmods it to the required
4879 mode, then resets the umask. All the relevant calls to fopen() in the
4880 content scanning code have been changed to use this function.
4882 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4883 to 24 hours. This avoids potential overflow problems when processing G
4884 and H retry rules. I suspect nobody ever tinkers with this value.
4886 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4888 PH/45 When the plaintext authenticator is running as a client, the server's
4889 challenges are checked to ensure they are valid base64 strings. By
4890 default, the authentication attempt is cancelled if an invalid string is
4891 received. Setting client_ignore_invalid_base64 true ignores these errors.
4892 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4893 they are received. Thus, the responses can be made to depend on the
4894 challenges. If an invalid string is ignored, an empty string is placed in
4897 PH/46 Messages that are created by the autoreply transport now contains a
4898 References: header, in accordance with RFCs 2822 and 3834.
4900 PH/47 Added authenticated_sender_force to the smtp transport.
4902 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4904 PH/49 Installed latest patch from the Sieve maintainer.
4906 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4907 was also set, the appendfile transport was unnecessarily scanning a
4908 directory of message files (e.g. for maildir delivery) to find the count
4909 of files (along with the size), even though it did not need this
4910 information. It now does the scan only if it needs to find either the
4911 size of the count of files.
4913 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4915 PH/52 Two bugs concerned with error handling when the smtp transport is
4918 (i) Exim was not creating retry information for temporary errors given
4919 for individual recipients after the DATA command when the smtp transport
4920 was used in LMTP mode. This meant that they could be retried too
4921 frequently, and not timed out correctly.
4923 (ii) Exim was setting the flag that allows error details to be returned
4924 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4925 recipients that were returned after the DATA command.
4927 PH/53 This is related to PH/52, but is more general: for any failing address,
4928 when detailed error information was permitted to be returned to the
4929 sender, but the error was temporary, then after the final timeout, only
4930 "retry timeout exceeded" was returned. Now it returns the full error as
4931 well as "retry timeout exceeded".
4933 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4934 do this, and (what is worse) MTAs that accept it.
4936 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4937 will now be deprecated.
4939 PH/56 New os.c-cygwin from the Cygwin maintainer.
4941 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4942 formats (previously only available in exim formats via -bpr, -bpru,
4943 and -bpra. Now also available in native and exiqgrep formats)
4945 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4946 with very large, slow to parse queues
4948 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4950 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4952 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4953 responses to authentication challenges, though it was showing the
4954 challenges; (ii) I've removed the CR characters from the debug output for
4957 PH/58 Allow for the insertion of a newline as well as a space when a string
4958 is turned into more than one encoded-word during RFC 2047 encoding. The
4959 Sieve code now uses this.
4961 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4962 data_4xx, lost_connection, tls_required.
4964 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4965 message was being sent as an SMTP response.
4967 PH/61 Add -l and -k options to exicyclog.
4969 PH/62 When verifying, if an address was redirected to one new address, so that
4970 verification continued, and the new address failed or deferred after
4971 having set something in $address_data, the value of $address_data was not
4972 passed back to the ACL. This was different to the case when no
4973 redirection occurred. The value is now passed back in both cases.
4975 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4976 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4977 use login_cap.h, so on its own it isn't the distinguishing feature. The
4978 new name refers directly to the setclassresources() function.
4980 PH/65 Added configuration files for NetBSD3.
4982 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4984 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4985 is preferred over IPv4.
4987 PH/68 The bounce_return_message and bounce_return_body options were not being
4988 honoured for bounces generated during the reception of non-SMTP messages.
4989 In particular, this applied to messages rejected by the ACL. This bug has
4990 been fixed. However, if bounce_return_message is true and bounce_return_
4991 body is false, the headers that are returned for a non-SMTP message
4992 include only those that have been read before the error was detected.
4993 (In the case of an ACL rejection, they have all been read.)
4995 PH/69 The HTML version of the specification is now built in a directory called
4996 spec_html instead of spec.html, because the latter looks like a path with
4997 a MIME-type, and this confuses some software.
4999 PH/70 Catch two compiler warnings in sieve.c.
5001 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
5002 function verify_get_ident() calls ip_connect() to connect a socket, but
5003 if the "connect()" function timed out, ip_connect() used to close the
5004 socket. However, verify_get_ident() also closes the socket later, and in
5005 between Exim writes to the log, which may get opened at this point. When
5006 the socket was closed in ip_connect(), the log could get the same file
5007 descriptor number as the socket. This naturally causes chaos. The fix is
5008 not to close the socket in ip_connect(); the socket should be closed by
5009 the function that creates it. There was only one place in the code where
5010 this was missing, in the iplookup router, which I don't think anybody now
5011 uses, but I've fixed it anyway.
5013 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
5014 well as to direct DNS lookups. Otherwise the handling of names in host
5015 lists is inconsistent and therefore confusing.
5021 PH/01 Two changes to the default runtime configuration:
5023 (1) Move the checks for relay_from_hosts and authenticated clients from
5024 after to before the (commented out) DNS black list checks.
5026 (2) Add control=submission to the relay_from_hosts and authenticated
5027 clients checks, on the grounds that messages accepted by these
5028 statements are most likely to be submissions.
5030 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
5032 (1) Generate an error if the third argument for the ${prvs expansion is
5035 (2) Treat a missing third argument of ${prvscheck as if it were an empty
5038 (3) Reset the variables that are obtained from the first argument of
5039 ${prvscheck and used in the second argument before leaving the code,
5040 because their memory is reclaimed, so using them afterwards may do
5043 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
5044 one (it's much easier than Tom thought :-).
5046 (5) Because of (4), we can now allow for the use of $prvscheck_result
5047 inside the third argument.
5049 PH/03 For some reason, the default setting of PATH when running a command from
5050 a pipe transport was just "/usr/bin". I have changed it to
5053 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
5054 anything to be listed in the output from -bV.
5056 PH/05 When a filter generated an autoreply, the entire To: header line was
5057 quoted in the delivery log line, like this:
5059 => >A.N.Other <ano@some.domain> <original@ddress> ...
5061 This has been changed so that it extracts the operative address. There
5062 may be more than one such address. If so, they are comma-separated, like
5065 => >ano@some.domain,ona@other.domain <original@ddress> ...
5067 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5068 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5069 not being looked up in the rDNS to get a host name, Exim was showing the
5070 IP address twice in Received: lines, even though the IP addresses were
5071 identical. For example:
5073 Received: from [1.2.3.4] (helo=[1.2.3.4])
5075 However, if the real host name was known, it was omitting the HELO data
5076 if it matched the actual IP address. This has been tidied up so that it
5077 doesn't show the same IP address twice.
5079 PH/07 When both +timestamp and +memory debugging was on, the value given by
5080 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5081 called by the debug printing, thereby overwriting the timestamp buffer.
5082 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5084 PH/08 When the original message was included in an autoreply transport, it
5085 always said "this is a copy of the message, including all the headers",
5086 even if body_only or headers_only was set. It now gives an appropriate
5089 PH/09 Applied a patch from the Sieve maintainer which:
5091 o fixes some comments
5092 o adds the (disabled) notify extension core
5093 o adds some debug output for the result of if/elsif tests
5094 o points to the current vacation draft in the documentation
5095 and documents the missing references header update
5099 o fixes a bug in processing the envelope test (when testing
5100 multiple envelope elements, the last element determined the
5103 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5104 Electronic Mail") by including:
5106 Auto-submitted: auto-generated
5108 in the messages that it generates (bounce messages and others, such as
5109 warnings). In the case of bounce messages for non-SMTP messages, there was
5110 also a typo: it was using "Auto_submitted" (underscore instead of
5111 hyphen). Since every message generated by Exim is necessarily in response
5112 to another message, thes have all been changed to:
5114 Auto-Submitted: auto-replied
5116 in accordance with these statements in the RFC:
5118 The auto-replied keyword:
5120 - SHOULD be used on messages sent in direct response to another
5121 message by an automatic process,
5123 - MUST NOT be used on manually-generated messages,
5125 - MAY be used on Delivery Status Notifications (DSNs) and Message
5126 Disposition Notifications (MDNs),
5128 - MUST NOT be used on messages generated by automatic or periodic
5129 processes, except for messages which are automatic responses to
5132 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5133 to the default Received: header definition.
5135 PH/12 Added log selector acl_warn_skipped (default on).
5137 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5138 variables because (a) they are in the wrong storage pool and (b) even if
5139 they were copied, it wouldn't work properly because of the caching.
5141 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5142 checking when decoding. Apparently there are clients that generate
5143 overlong encoded strings. Why am I not surprised?
5145 PH/15 If the first argument of "${if match_address" was not empty, but did not
5146 contain an "@" character, Exim crashed. Now it writes a panic log message
5147 and treats the condition as false.
5149 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5151 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5152 extension "envelope-auth". The code is finished and in agreement with
5153 other implementations, but there is no documentation so far and in fact,
5154 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5155 not changing the active code.
5157 Print executed "if" and "elsif" statements when debugging is used. This
5158 helps a great deal to understand what a filter does.
5160 Document more things not specified clearly in RFC3028. I had all this
5161 sorted out, when out of a sudden new issues came to my mind. Oops."
5163 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5166 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5167 canonical form (as documented). However, after a host name lookup from
5168 the IP address, check_host() was doing a simple string comparison with
5169 addresses acquired from the DNS when checking that the found name did
5170 have the original IP as one of its addresses. Since any found IPv6
5171 addresses are likely to be in abbreviated form, the comparison could
5172 fail. Luckily, there already exists a function for doing the comparison
5173 by converting both addresses to binary, so now that is used instead of
5174 the text comparison.
5176 PH/20 There was another similar case to PH/19, when a complete host name was
5177 given in a host list; looking up its IP address could give an abbreviated
5178 form, whereas the current host's name might or might not be abbreviated.
5179 The same fix has been applied.
5185 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5186 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5189 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5190 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5192 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5194 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5195 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5196 Also, for "Auto-Submitted", treat anything other than "no" as
5197 non-personal, in accordance with RFC 3834. (Previously it treated
5198 anything starting "auto-" as non-personal.)
5200 TF/01 The control=submission/name=... option had a problem with syntax
5201 errors if the name included a slash character. The /name= option
5202 now slurps the rest of the string, so it can include any characters
5203 but it must come last in the list of options (after /sender_retain
5206 PH/05 Some modifications to the interface to the fake nameserver for the new
5214 TK/01 Added the "success_on_redirect" address verification option. See
5215 NewStuff for rationale and an example.
5217 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5219 PH/02 Patch to exigrep to allow it to work on syslog lines.
5221 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5222 fread() to skip over the body file's header line, because in Cygwin the
5223 header line is locked and is inaccessible.
5225 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5226 co-exist for some time) to make it clear that it is the Exim ID that is
5227 referenced, not the Message-ID: header line.
5229 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5230 string_format() function, because snprintf() does not exist on all
5233 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5234 ${stat: expansion item.
5236 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5237 protocol synchronization error", to keep the pedants happy.
5239 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5240 well as for IRIX systems, when gcc is being used. See the host.c source
5243 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5245 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5248 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5249 transport and to the smtp transport in LMTP mode.
5251 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5253 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5254 run a filter in a subprocess. This could lead to confusion in subsequent
5255 lookups in the parent process. There should also be a search_tidyup() at
5256 the end of the subprocess.
5258 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5259 only if the host matched helo_try_verify_hosts, which caused the
5260 verification to occur when the EHLO/HELO command was issued. The ACL just
5261 tested the remembered result. Now, if a previous verification attempt has
5262 not happened, "verify = helo" does it there and then.
5264 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5266 TK/03 Fix log output including CR from clamd.
5268 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5269 exist provoked a memory error which could cause a segfault.
5271 PH/15 Installed PCRE 6.2
5273 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5275 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5276 of the problem. Specifically, suggested +O2 rather than +O1 for the
5279 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5281 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5282 record, the port was not correctly logged when the outgoing_port log
5283 selector was set (it logged the transort's default port).
5285 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5286 fallback_hosts, and "hosts" in the smtp transport.
5288 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5289 host errors such as "Connection refused".
5291 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5292 authentication with radiusclient 0.4.9:
5294 - Error returned from rc_read_config was caught wrongly
5295 - Username/password not passed on to radius server due to wrong length.
5297 The presumption is that some radiusclient API changes for 4.51/PH/17
5298 were not taken care of correctly. The code is still untested by me (my
5299 Linux distribution still has 0.3.2 of radiusclient), but it was
5300 contributed by a Radius user.
5302 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5303 expanding the "port" option of the smtp transport.
5305 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5306 while reading a MIME header. Thanks to Tom Hughes for a patch.
5308 PH/24 Include config.h inside local_scan.h so that configuration settings are
5311 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5312 that in an ACL for RCPT (for example), you can examine exactly what was
5315 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5316 commands, but it was not correctly comparing the address with the actual
5317 client host address. Thus, it would show the EHLO address in Received:
5318 header lines when this was not necessary.
5320 PH/27 Added the % operator to ${eval:}.
5322 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5323 it should be ignoring failures (because with -C, for example, it has lost
5324 privilege). It wasn't ignoring creation failures other than "already
5327 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5330 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5331 because an input line was too long, either on its own, or by virtue of
5332 too many continuations, the temporary file was not being removed, and the
5333 return code was incorrect.
5335 PH/31 Missing "BOOL" in function definition in filtertest.c.
5337 PH/32 Applied Sieve patches from the maintainer.
5339 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5341 PH/33 Added "verify = not_blind".
5343 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5344 Local/Makefile (with some defaults set). These are used in built scripts
5345 such as exicyclog, but they have never been used in the exim_install
5346 script (though there are many overriding facilities there). I have
5347 arranged that the exim_install script now takes note of these two
5350 PH/35 Installed configuration files for Dragonfly.
5352 PH/36 When a locally submitted message by a trusted user did not contain a
5353 From: header, and the sender address was obtained from -f or from an SMTP
5354 MAIL command, and the trusted user did not use -F to supply a sender
5355 name, $originator_name was incorrectly used when constructing a From:
5356 header. Furthermore, $originator_name was used for submission mode
5357 messages from external hosts without From: headers in a similar way,
5358 which is clearly wrong.
5360 PH/37 Added control=suppress_local_fixups.
5362 PH/38 When log_selector = +received_sender was set, and the addition of the
5363 sender made the log line's construction buffer exactly full, or one byte
5364 less than full, an overflow happened when the terminating "\n" was
5367 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5368 when the result of a list match is failure because a DNS lookup failed.
5370 PH/40 RM_COMMAND is now used in the building process.
5372 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5373 the "build-* directories that it finds.
5375 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5376 address in a domain literal was a prefix of an interface address.
5378 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5379 when verifying a sender address, unless rewrite_headers is false.
5381 PH/44 Wrote a long comment about why errors_to addresses are verified as
5382 recipients, not senders.
5384 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5385 the ratelimit ACL was added.
5387 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5389 PH/46 Added extra information about PostgreSQL errors to the error string.
5391 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5392 avoiding the need to install special zones in a real server. This is
5393 backwards compatible; if it can't find the fake resolver, it drops back.
5394 Thus, both old and new test suites can be run.
5396 TF/02 Added util/ratelimit.pl
5398 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5399 clock is set back in time.
5401 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5402 Candler <B.Candler@pobox.com>.
5404 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5405 OK for addresses that are the result of redirections.
5407 PH/48 A number of further additions for the benefit of the new test suite,
5408 including a fake gethostbyname() that interfaces to the fake DNS resolver
5411 TF/06 The fix for widen_domains has also been applied to qualify_single and
5412 search_parents which are the other dnslookup options that can cause
5415 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5418 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5420 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5421 service" if it encountered an MX record with an empty target hostname.
5422 The message is now "an MX or SRV record indicated no SMTP service".
5424 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5425 if the DNS of the sending site is misconfigured. This is quite a
5426 common situation. This change restores the behaviour of treating a
5427 helo verification defer as a failure.
5429 PH/51 If self=fail was set on a router, the bounce message did not include the
5430 actual error message.
5436 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5438 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5439 command itself ended in error, the underlying message about the transport
5440 filter timeout was being overwritten with the pipe command error. Now the
5441 underlying error message should be appended to the second error message.
5443 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5445 PH/02 Reduce the amount of output that "make" produces by default. Full output
5446 can still be requested.
5448 PH/03 The warning log line about a condition test deferring for a "warn" verb
5449 was being output only once per connection, rather than after each
5450 occurrence (because it was using the same function as for successful
5451 "warn" verbs). This seems wrong, so I have changed it.
5453 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5454 it should not have, which might have caused a crash in the right
5455 circumstances, but probably never did.
5457 PH/04 Installed a modified version of Tony Finch's patch to make submission
5458 mode fix the return path as well as the Sender: header line, and to
5459 add a /name= option so that you can make the user's friendly name appear
5462 TF/03 Added the control = fakedefer ACL modifier.
5464 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5465 Mark Lowes for thorough testing.
5467 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5469 TK/03 Merged latest SRS patch from Miles Wilton.
5471 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5472 with the definition in sysexits.h (which is #included earlier).
5473 Fortunately, Exim does not actually use EX_OK. The code used to try to
5474 preserve the sysexits.h value, by assuming that macro definitions were
5475 scanned for macro replacements. I have been disabused of this notion,
5476 so now the code just undefines EX_OK before #including unistd.h.
5478 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5479 in the smtp transport. When a block could not be written in a single
5480 write() function, the timeout was being re-applied to each part-write.
5481 This seems wrong - if the receiver was accepting one byte at a time it
5482 would take for ever. The timeout is now adjusted when this happens. It
5483 doesn't have to be particularly precise.
5485 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5486 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5488 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5489 without a domain if the check to <postmaster@domain> fails.
5491 SC/01 Eximstats: added -xls and the ability to specify output files
5492 (patch written by Frank Heydlauf).
5494 SC/02 Eximstats: use FileHandles for outputting results.
5496 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5498 SC/04 Eximstats: fixed display of large numbers with -nvr option
5500 SC/05 Eximstats: fixed merging of reports with empty tables.
5502 SC/06 Eximstats: added the -include_original_destination flag
5504 SC/07 Eximstats: removed tabs and trailing whitespace.
5506 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5508 TK/06 MBOX spool code: Add real "From " MBOX separator line
5509 so the .eml file is really in mbox format (even though
5510 most programs do not really care). Patch from Alex Miller.
5512 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5513 The latter is generated from $received_to and is only set if the
5514 message has one envelope recipient. SA can use these headers,
5515 obviously out-of-the-box. Patch from Alex Miller.
5517 PH/08 The ${def test on a variable was returning false if the variable's
5518 value was "0", contrary to what the specification has always said!
5519 The result should be true unless the variable is empty.
5521 PH/09 The syntax error of a character other than { following "${if
5522 def:variable_name" (after optional whitespace) was not being diagnosed.
5523 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5524 accidental colon was present, for example, could give incorrect results.
5526 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5527 result is used (not including appendfile, where other changes are about
5530 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5531 This involved changing a lot of size variables from int to off_t. It
5532 should work with maildirs and everything.
5534 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5535 spamd dying while we are connected to it.
5537 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5540 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5541 function for BDB 4.3.
5543 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5545 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5546 the sources for PCRE have been re-arranged and more files are now
5549 PH/15 The code I had for printing potentially long long variables in PH/11
5550 above was not the best (it lost precision). The length of off_t variables
5551 is now inspected at build time, and an appropriate printing format (%ld
5552 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5553 to be "long long int" or "long int". This is needed for the internal
5554 formatting function string_vformat().
5556 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5557 the configuration file to be ":syslog", then the script "guesses" where
5558 the logs files are, rather than using the compiled in default. In our
5559 case the guess is not the same as the compiled default, so the script
5560 suddenly stopped working when I started to use syslog. The patch checks
5561 to see if log_file_path is "". If so, it attempts to read it from exim
5562 with no configuration file to get the compiled in version, before it
5563 falls back to the previous guessing code."
5565 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5566 implementing BATV in an Exim configuration. See NewStuff for the gory
5569 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5570 Makefile that are specific to HP-UX.
5572 PH/18 If the "use_postmaster" option was set for a recipient callout together
5573 with the "random" option, the postmaster address was used as the MAIL
5574 FROM address for the random test, but not for the subsequent recipient
5575 test. It is now used for both.
5577 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5578 patch removes a few documentation additions to RFC 3028, because the
5579 latest draft now contains them. It adds the new en;ascii-case comparator
5580 and a new error check for 8bit text in MIME parts. Comparator and
5581 require names are now matched exactly. I enabled the subaddress
5582 extension, but it is not well tested yet (read: it works for me)."
5584 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5585 rework some of the code of TK/09 above to avoid the hardwired use of
5586 "%lld" and "long long". Replaced the call to snprintf() with a call to
5589 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5590 records point to non-existent hosts", "retry timeout exceeded", and
5591 "retry time not reached for any host after a long failure period".
5593 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5594 experimental DomainKeys support:
5596 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5597 (2) On an error such as an illegally used "control", the wrong name for
5598 the control was given.
5600 These problems did NOT occur unless DomainKeys support was compiled.
5602 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5604 PH/24 Added ${if match_ip condition.
5606 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5607 databases so that it will be absolutely obvious if a crash occurs in the
5608 DB library. This is a regular occurrence (often caused by mis-matched
5611 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5612 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5613 picked up on a user's system that detects such things. There doesn't seem
5614 to be a gcc warning option for this - only an attribute that has to be
5615 put on the function's prototype. It seems that in Fedora Core 4 they have
5616 set this on a number of new functions. No doubt there will be more in due
5619 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5620 specify a transport. However, if an address that was verified by such a
5621 router was the subject of a callout, Exim crashed because it tried to
5622 read the rcpt_include_affixes from the non-existent transport. Now it
5623 just assumes that the setting of that option is false. This bug was
5624 introduced by 4.51/PH/31.
5626 PH/28 Changed -d+all to exclude +memory, because that information is very
5627 rarely of interest, but it makes the output a lot bigger. People tend to
5628 do -d+all out of habit.
5630 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5631 code in os-type was giving problems when libc.so lives in lib64, like on
5634 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5635 aren't the modern standard, and it seems that some systems' include files
5636 don't always have them. Exim was already checking for some of the newer
5637 ones like T_AAAA, and defining it itself. I've added checks for all the
5638 record types that Exim uses.
5640 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5641 not automatically generating a new one, as it is supposed to. This
5642 prevented TLS from working. If the file did exist, but contained invalid
5643 data, a new version was generated, as expected. It was only the case of a
5644 non-existent file that was broken.
5646 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5647 with a change in libdomainkeys > 0.64.
5649 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5650 from DNS. If the selector record carries the flag, it now has
5651 precedence over the domain-wide flag.
5653 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5655 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5656 the use of a port name that isn't defined in /etc/services) occurred, the
5657 message was deferred as in a normal delivery, and thus remained on the
5658 spool, instead of being failed because of the mua_wrapper setting. This
5659 is now fixed, and I tidied up some of the mua_wrapper messages at the
5662 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5663 the messages in a hash of arrays rather than using individual hashes.
5664 This is a bit cleaner and results in dramatic memory savings, albeit
5665 at a slight CPU cost.
5667 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5668 as requested by Marc Sherman.
5670 SC/10 Eximstats: added histograms for user specified patterns as requested
5673 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5675 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5676 fopen() in the content-scanning modules that did not already have it.
5682 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5683 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5685 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5687 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5689 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5690 to test Sieve filters that use "vacation".
5692 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5693 that changes the way the GnuTLS parameters are stored in the cache file.
5694 The new format can be generated externally. For backward compatibility,
5695 if the data in the cache doesn't make sense, Exim assumes it has read an
5696 old-format file, and it generates new data and writes a new file. This
5697 means that you can't go back to an older release without removing the
5700 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5701 work if there are any delivery delays because "one_time" forces the
5702 parent to be marked "delivered", so its unseen clone is never tried
5703 again. For this reason, Exim now forbids the simultaneous setting of
5706 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5707 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5708 entry if you want to know the details. The fix, however, neglected to
5709 consider the case when local delivery batching is involved. The test for
5710 "previously delivered" was not happening when checking to see if an
5711 address could be batched with a previous (undelivered) one; under
5712 certain circumstances this could lead to multiple deliveries to the same
5715 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5716 in its include files, and this causes problems building Exim.
5718 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5719 header_syntax) but Exim was just ignoring anything given after a slash.
5720 In particular, this caused confusion with an attempt to use "verify =
5721 reverse_host_lookup/defer_ok". An error is now given when options are
5722 supplied for verify items that do not have them. (Maybe reverse_host_
5723 lookup should have a defer_ok option, but that's a different point.)
5725 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5726 defined by RFC 821) to 2048, because there were problems with some AUTH
5727 commands, and RFC 1869 says the size should be increased for extended
5728 SMTP commands that take arguments.
5730 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5733 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5734 "unknown" error; now it says that the functionality isn't in the binary.
5736 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5737 an address' error message when a string expansion fails (syntax or
5738 whatever). Otherwise the password may appear in the log. Following change
5739 PH/42 below, there is no longer a chance of it appearing in a bounce
5742 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5744 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5745 of its addresses was ever tried. (Bugzilla bug #2.)
5747 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5748 the result incorrectly in the debug output. (It correctly added a newline
5749 to what was transported.)
5751 TF/01 Added $received_time.
5753 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5754 commented out examples of how to interface to a virus scanner and to
5755 SpamAssassin. Also added commented examples of av_scanner and
5756 spamd_address settings.
5758 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5759 and controls are allowed in which ACLs. There were a couple of minor
5760 errors. Some of the entries in the conditions table (which is a table of
5761 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5762 negation of where the condition IS allowed.
5764 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5766 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5767 header file does not have a version number, so I've had to invent a new
5768 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5769 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5770 radiusclient), but it was contributed by a Radius user.
5772 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5773 files or directories, for OpenSSL.
5775 PH/19 When an Exim process that is running as root has to create an Exim log
5776 file, it does so in a subprocess that runs as exim:exim so as to get the
5777 ownership right at creation (otherwise, other Exim processes might see
5778 the file with the wrong ownership). There was no test for failure of this
5779 fork() call, which would lead to the process getting stuck as it waited
5780 for a non-existent subprocess. Forks do occasionally fail when resources
5781 run out. I reviewed all the other calls to fork(); they all seem to check
5784 PH/20 When checking for unexpected SMTP input at connect time (before writing
5785 the banner), Exim was not dealing correctly with a non-positive return
5786 from the read() function. If the client had disconnected by this time,
5787 the result was a log entry for a synchronization error with an empty
5788 string after "input=" when read() returned zero. If read() returned -1
5789 (an event I could not check), uninitialized data bytes were printed.
5790 There were reports of junk text (parts of files, etc) appearing after
5793 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5795 PH/22 Added support for macro redefinition, and (re)definition in between
5796 driver and ACL definitions.
5798 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5799 forgetting to use the resulting value; it was using the unexpanded value.
5801 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5802 hadn't been configured. The fix is from Juergen Kreileder, who
5803 understands it better than I do:
5805 "Here's what I see happening with three configured cyrus_sasl
5806 authenticators configured (plain, login, cram-md5):
5808 On startup auth_cyrus_sasl_init() gets called for each of these.
5809 This means three calls to sasl_listmech() without a specified mech_list.
5810 => SASL tests which mechs of all available mechs actually work
5811 => three warnings about OTP not working
5812 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5814 With the patch, sasl_listmech() also gets called three times. But now
5815 SASL's mech_list option is set to the server_mech specified in the the
5816 authenticator. Or in other words, the answer from sasl_listmech()
5817 gets limited to just the mech you're testing for (which is different
5819 => the return list contains just 'plain' or 'login', 'cram-md5' or
5820 nothing depending on the value of ob->server_mech.
5822 I've just tested the patch: Authentication still works fine,
5823 unavailable mechs specified in the exim configuration are still
5824 caught, and the auth.log warnings about OTP are gone."
5826 PH/25 When debugging is enabled, the contents of the command line are added
5827 to the debugging output, even when log_selector=+arguments is not
5830 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5831 answer is "GNU", and only if the return is "GNU/something" is the answer
5834 PH/27 $acl_verify_message is now set immediately after the failure of a
5835 verification in an ACL, and so is available in subsequent modifiers. In
5836 particular, the message can be preserved by coding like this:
5838 warn !verify = sender
5839 set acl_m0 = $acl_verify_message
5841 Previously, $acl_verify_message was set only while expanding "message"
5842 and "log_message" when a very denied access.
5844 PH/28 Modified OS/os.c-Linux with
5846 -#ifndef OS_LOAD_AVERAGE
5847 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5849 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5850 nomenclature these days.)
5852 PH/29 Installed patch from the Sieve maintainer that adds the options
5853 sieve_useraddress and sieve_subaddress to the redirect router.
5855 PH/30 In these circumstances:
5856 . Two addresses routed to the same list of hosts;
5857 . First host does not offer TLS;
5858 . First host accepts first address;
5859 . First host gives temporary error to second address;
5860 . Second host offers TLS and a TLS session is established;
5861 . Second host accepts second address.
5862 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5863 and peerdn, if requested) that were in fact used only for the second
5866 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5867 attention to any local part prefix or suffix that was matched by the
5868 router that accepted the address. It now behaves in the same way as it
5869 does for delivery: the affixes are removed from the local part unless
5870 rcpt_include_affixes is set on the transport.
5872 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5873 timeout during the DATA phase of an incoming message.
5875 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5876 applied a patch sanctioned by the Sieve maintainer.
5878 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5879 the uid or gid is negative. A case of a negative gid caused this to be
5880 noticed. The fix allows for either to be negative.
5882 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5883 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5886 PH/36 The change PH/12 above was broken. Fixed it.
5888 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5889 the grounds that routing the same address twice would always produce the
5890 same answer. This might have been true once, but it is certainly no
5891 longer true now. Routing a child address may depend on the previous
5892 routing that produced that child. Some complicated redirection strategies
5893 went wrong when messages had multiple recipients, and made Exim's
5894 behaviour dependent on the order in which the addresses were given.
5896 I have moved the duplicate checking until after the routing is complete.
5897 Exim scans the addresses that are assigned to local and remote
5898 transports, and removes any duplicates. This means that more work will be
5899 done, as duplicates will always all be routed, but duplicates are
5900 presumably rare, so I don't expect this is of any significance.
5902 For deliveries to pipes, files, and autoreplies, the duplicate checking
5903 still happens during the routing process, since they are not going to be
5906 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5907 It corrects a timeout issue with spamd. This is Ian's comment: "The
5908 background is that sometimes spamd either never reads data from a
5909 connection it has accepted, or it never writes response data. The exiscan
5910 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5911 blindly assumes that writes won't block so it may never time out."
5913 PH/39 Allow G after quota size as well as K and M.
5915 PH/40 The value set for $authenticated_id in an authenticator may not contain
5916 binary zeroes or newlines because the value is written to log lines and
5917 to spool files. There was no check on this. Now the value is run through
5918 the string_printing() function so that such characters are converted to
5919 printable escape sequences.
5921 PH/41 $message_linecount is a new variable that contains the total number of
5922 lines in the message. Compare $body_linecount, which is the count for the
5925 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5926 bounce and delay warning messages, except in certain special cases, which
5929 (a) An SMTP error message from a remote host;
5930 (b) A message specified in a :fail: redirection;
5931 (c) A message specified in a "fail" command in a system filter;
5932 (d) A message specified in a FAIL return from the queryprogram router;
5933 (e) A message specified by the cannot_route_message router option.
5935 In these cases only, Exim does include the error details in bounce and
5936 warning messages. There are also a few cases where bland messages such
5937 as "unrouteable address" or "local delivery error" are given.
5939 PH/43 $value is now also set for the "else" part of a ${run expansion.
5941 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5942 being worked on, but at least Exim now implements the latest version to
5945 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5946 process to complete was treated as a delivery failure, a timeout while
5947 writing the message to the pipe was logged, but erroneously treated as a
5948 successful delivery. Such timeouts include transport filter timeouts. For
5949 consistency with the overall process timeout, these timeouts are now
5950 treated as errors, giving rise to delivery failures by default. However,
5951 there is now a new Boolean option for the pipe transport called
5952 timeout_defer, which, if set TRUE, converts the failures into defers for
5953 both kinds of timeout. A transport filter timeout is now identified in
5956 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5957 systems where "make" and "gmake" are different, calling "gmake" at top
5958 level broke things. I've arranged for the value of $(MAKE) to be passed
5959 from the Makefile to this script so that it can call the same version of
5963 A note about Exim versions 4.44 and 4.50
5964 ----------------------------------------
5966 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5967 changes of various kinds. As a consequence, a big documentation update was
5968 needed. This delayed the release for rather longer than seemed good, especially
5969 in the light of a couple of (minor) security issues. Therefore, the changes
5970 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5971 release. So 4.44 and 4.50 are in effect two different branches that both start
5974 I have left the 4.50 change log unchanged; it contains all the changes since
5975 4.43. The change log for 4.44 is below; many of its items are identical to
5976 those for 4.50. This seems to be the most sensible way to preserve the
5977 historical information.
5983 1. Minor wording change to the doc/README.SIEVE file.
5985 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5986 computation of the current number of files was incorrect.
5988 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5989 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5992 4. Give more explanation in the error message when the command for a transport
5993 filter fails to execute.
5995 5. There are several places where Exim runs a non-Exim command in a
5996 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5997 was being done only for the command run by the queryprogram router. It is
5998 now done for all such subprocesses. The other cases are: ${run, transport
5999 filters, and the commands run by the lmtp and pipe transports.
6001 6. Added CONFIGURE_GROUP build-time option.
6003 7. Some older OS have a limit of 256 on the maximum number of file
6004 descriptors. Exim was using setrlimit() to set 1000 as a large value
6005 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6006 systems. I've change it so that if it can't get 1000, it tries for 256.
6008 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6009 was an oversight, and furthermore, ever since the addition of extra
6010 controls (e.g. 4.43/32), the checks on when to allow different forms of
6011 "control" were broken. There should now be diagnostics for all cases when a
6012 control that does not make sense is encountered.
6014 9. Added the /retain_sender option to "control=submission".
6016 10. $recipients is now available in the predata ACL (oversight).
6018 11. Tidy the search cache before the fork to do a delivery from a message
6019 received from the command line. Otherwise the child will trigger a lookup
6020 failure and thereby defer the delivery if it tries to use (for example) a
6021 cached ldap connection that the parent has called unbind on.
6023 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6024 of $address_data from the recipient verification was clobbered by the
6025 sender verification.
6027 13. The value of address_data from a sender verification is now available in
6028 $sender_address_data in subsequent conditions in the ACL statement.
6030 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
6032 15. Added a new option "connect=<time>" to callout options, to set a different
6035 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6036 was its contents. (It was OK if the option was not defined at all.)
6038 17. A "Completed" log line is now written for messages that are removed from
6039 the spool by the -Mrm option.
6041 18. New variables $sender_verify_failure and $recipient_verify_failure contain
6042 information about exactly what failed.
6044 19. Added -dd to debug only the daemon process.
6046 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
6047 handles timeouts, both on the server side and network timeouts. Renamed the
6048 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
6050 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
6051 It is now set to "smtps".
6053 22. $host_address is now set to the target address during the checking of
6054 ignore_target_hosts.
6056 23. When checking ignore_target_hosts for an ipliteral router, no host name was
6057 being passed; this would have caused $sender_host_name to have been used if
6058 matching the list had actually called for a host name (not very likely,
6059 since this list is usually IP addresses). A host name is now passed as
6062 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6063 code that specifies a non-restarting handler (typically sigaction() in
6064 modern systems) in an attempt to fix a rare and obscure crash bug.
6066 25. Narrowed the window for a race in the daemon that could cause it to ignore
6067 SIGCHLD signals. This is not a major problem, because they are used only to
6068 wake it up if nothing else does.
6070 26. A malformed maildirsize file could cause Exim to calculate negative values
6071 for the mailbox size or file count. Odd effects could occur as a result.
6072 The maildirsize information is now recalculated if the size or filecount
6075 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6076 support for a long time. Removed HAVE_SYS_VFS_H.
6078 28. Installed the latest version of exipick from John Jetmore.
6080 29. In an address list, if the pattern was not a regular expression, an empty
6081 subject address (from a bounce message) matched only if the pattern was an
6082 empty string. Non-empty patterns were not even tested. This was the wrong
6083 because it is perfectly reasonable to use an empty address as part of a
6084 database query. An empty address is now tested by patterns that are
6085 lookups. However, all the other forms of pattern expect the subject to
6086 contain a local part and a domain, and therefore, for them, an empty
6087 address still always fails if the pattern is not itself empty.
6089 30. Exim went into a mad DNS loop when attempting to do a callout where the
6090 host was specified on an smtp transport, and looking it up yielded more
6091 than one IP address.
6093 31. Re-factored the code for checking spool and log partition space into a
6094 function that finds that data and another that does the check. The former
6095 is then used to implement four new variables: $spool_space, $log_space,
6096 $spool_inodes, and $log_inodes.
6098 32. The RFC2047 encoding function was originally intended for short strings
6099 such as real names; it was not keeping to the 75-character limit for
6100 encoded words that the RFC imposes. It now respects the limit, and
6101 generates multiple encoded words if necessary. To be on the safe side, I
6102 have increased the buffer size for the ${rfc2047: expansion operator from
6105 33. It is now permitted to omit both strings after an "if" condition; if the
6106 condition is true, the result is "true". As before, when the second string
6107 is omitted, a false condition yields an empty string. This makes it less
6108 cumbersome to write custom ACL and router conditions.
6110 34. Failure to deliver a bounce message always caused it to be frozen, even if
6111 there was an errors_to setting on the router. The errors_to setting is now
6114 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6115 canonical form (fully expanded) before being placed in
6116 $sender_host_address.
6118 36. The table in the code that translates DNS record types into text (T_A to
6119 "A" for instance) was missing entries for NS and CNAME. It is just possible
6120 that this could have caused confusion if both these types were looked up
6121 for the same domain, because the text type is used as part of Exim's
6122 per-process caching. But the chance of anyone hitting this buglet seems
6125 37. The dnsdb lookup has been extended in a number of ways.
6127 (1) There is a new type, "zns", which walks up the domain tree until it
6128 finds some nameserver records. It should be used with care.
6130 (2) There is a new type, "mxh", which is like "mx" except that it returns
6131 just the host names, not the priorities.
6133 (3) It is now possible to give a list of domains (or IP addresses) to be
6134 looked up. The behaviour when one of the lookups defers can be
6135 controlled by a keyword.
6137 (4) It is now possible to specify the separator character for use when
6138 multiple records are returned.
6140 38. The dnslists ACL condition has been extended: it is now possible to supply
6141 a list of IP addresses and/or domains to be looked up in a particular DNS
6144 39. Added log_selector=+queue_time_overall.
6146 40. When running the queue in the test harness, wait just a tad after forking a
6147 delivery process, to get repeatability of debugging output.
6149 41. Include certificate and key file names in error message when GnuTLS fails
6150 to set them up, because the GnuTLS error message doesn't include the name
6151 of the failing file when there is a problem reading it.
6153 42. Allow both -bf and -bF in the same test run.
6155 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6157 44. The "Exiscan patch" is now merged into the mainline Exim source.
6159 45. Sometimes the final signoff response after QUIT could fail to get
6160 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6161 < 0 before doing a fflush(). This bug looks as though it goes back to the
6162 introduction of TLS in release 3.20, but "sometimes" must have been rare
6163 because the tests only now provoked it.
6165 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6166 (this can affect the format of dates).
6168 47. exim_tidydb, when checking for the continued existence of a message for
6169 which it has found a message-specific retry record, was not finding
6170 messages that were in split spool directories. Consequently, it was
6171 deleting retry records that should have stayed in existence.
6173 48. Steve fixed some bugs in eximstats.
6175 49. The SPA authentication driver was not abandoning authentication and moving
6176 on to the next authenticator when an expansion was forced to fail,
6177 contradicting the general specification for all authenticators. Instead it
6178 was generating a temporary error. It now behaves as specified.
6180 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6181 (the order specifies the preference for clients). The order is now AES256,
6182 AES128, 3DES, ARCFOUR128.
6184 51. Small patch to Sieve code - explicitly set From: when generating an
6187 52. Exim crashed if a remote delivery caused a very long error message to be
6188 recorded - for instance if somebody sent an entire SpamAssassin report back
6189 as a large number of 550 error lines. This bug was coincidentally fixed by
6190 increasing the size of one of Exim's internal buffers (big_buffer) that
6191 happened as part of the Exiscan merge. However, to be on the safe side, I
6192 have made the code more robust (and fixed the comments that describe what
6195 53. Now that there can be additional text after "Completed" in log lines (if
6196 the queue_time_overall log selector is set), a one-byte patch to exigrep
6197 was needed to allow it to recognize "Completed" as not the last thing in
6200 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6201 patch that reportedly fixes this has been added. I am not expert enough to
6202 create a test for it. This is what the patch creator wrote:
6204 "I found a little strange behaviour of ldap code when working with
6205 Windows 2003 AD Domain, where users was placed in more than one
6206 Organization Units. When I tried to give exim partial DN, the exit code
6207 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6208 But simultaneously result of request was absolutely normal ldap result,
6209 so I produce this patch..."
6211 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6212 REFERENCE, so I have modified the code to exclude the patch when that macro
6215 55. Some experimental protocols are using DNS PTR records for new purposes. The
6216 keys for these records are domain names, not reversed IP addresses. The
6217 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6218 leaves it alone. Component reversal etc. now happens only for IP addresses.
6221 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6223 57. Double the size of the debug message buffer (to 2048) so that more of very
6224 long debug lines gets shown.
6226 58. The exicyclog utility now does better if the number of log files to keep
6227 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6229 59. Two changes related to the smtp_active_hostname option:
6231 (1) $smtp_active_hostname is now available as a variable.
6232 (2) The default for smtp_banner uses $smtp_active_hostname instead
6233 of $primary_hostname.
6235 60. The host_aton() function is supposed to be passed a string that is known
6236 to be a valid IP address. However, in the case of IPv6 addresses, it was
6237 not checking this. This is a hostage to fortune. Exim now panics and dies
6238 if the condition is not met. A case was found where this could be provoked
6239 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6240 components; fortuitously, this particular loophole had already been fixed
6241 by change 4.50/55 above.
6243 If there are any other similar loopholes, the new check in host_aton()
6244 itself should stop them being exploited. The report I received stated that
6245 data on the command line could provoke the exploit when Exim was running as
6246 exim, but did not say which command line option was involved. All I could
6247 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6248 running as the user.
6251 61. There was a buffer overflow vulnerability in the SPA authentication code
6252 (which came originally from the Samba project). I have added a test to the
6253 spa_base64_to_bits() function which I hope fixes it.
6256 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6257 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6259 63. The daemon start-up calls getloadavg() while still root for those OS that
6260 need the first call to be done as root, but it missed one case: when
6261 deliver_queue_load_max is set with deliver_drop_privilege. This is
6262 necessary for the benefit of the queue runner, because there is no re-exec
6263 when deliver_drop_privilege is set.
6265 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6266 This has been fixed.
6268 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6269 was in use, was not putting the data itself into the right store pool;
6270 consequently, it could be overwritten for a subsequent message in the same
6271 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6274 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6276 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6277 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6278 the function were treating the return as a boolean value, which happened to
6279 work because 0=false and not-0=true, but is not correct code.
6281 68. The host_aton() function was not handling scoped IPv6 addresses (those
6282 with, for example, "%eth0" on the end) correctly.
6284 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6285 negated items (that is, ~something) in unsigned ints. Some compilers
6286 apparently mutter when there is no cast.
6288 70. If an address verification called from an ACL failed, and did not produce a
6289 user-specific message (i.e. there was only a "system" message), nothing was
6290 put in $acl_verify_message. In this situation, it now puts the system
6293 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6294 session; change 4.31/43 added the unwanted input to the log line - except
6295 that it did not do this in the start of session case. It now does.
6297 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6298 This is wrong and can cause the other end to generate a synchronization
6299 error if it is another Exim or anything else that does the synchronization
6300 check. A QUIT command is no longer sent after a timeout.
6302 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6303 during host lookups.
6305 74. The defer_ok option of callout verification was not working if it was used
6306 when verifying addresses in header lines, that is, for this case:
6308 verify = header_sender/callout=defer_ok
6310 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6311 those file descriptors could be used for SMTP connections. If anything
6312 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6313 could be sent to the SMTP client, causing chaos. The daemon now opens
6314 stdin, stdout, and stderr to /dev/null when it puts itself into the
6317 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6318 log by default. The user can override this with suitable Perl magic.
6320 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6321 the log message when discard triggers, was not working for the DATA ACL or
6322 for the non-SMTP ACL.
6324 78. Error message wording change in sieve.c.
6326 79. If smtp_accept_max_per_host was set, the number of connections could be
6327 restricted to fewer than expected, because the daemon was trying to set up
6328 a new connection before checking whether the processes handling previous
6329 connections had finished. The check for completed processes is now done
6330 earlier. On busy systems, this bug wouldn't be noticed because something
6331 else would have woken the daemon, and it would have reaped the completed
6334 80. If a message was submitted locally by a user whose login name contained one
6335 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6336 It caused a spool format error. I have fixed the spool reading code. A
6337 related problem was that the "from" clause in the Received: line became
6338 illegal because of the space(s). It is now covered by ${quote_local_part.
6340 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6343 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6345 83. Patch from Sieve maintainer to fix unterminated string problem in
6346 "vacation" handling.
6348 84. Some minor changes to the Linux configuration files to help with other
6349 OS variants using glibc.
6351 85. One more patch for Sieve to update vacation handling to latest spec.
6354 ----------------------------------------------------
6355 See the note above about the 4.44 and 4.50 releases.
6356 ----------------------------------------------------
6362 1. Change 4.43/35 introduced a bug that caused file counts to be
6363 incorrectly computed when quota_filecount was set in an appendfile
6366 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6367 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6370 3. Give more explanation in the error message when the command for a transport
6371 filter fails to execute.
6373 4. There are several places where Exim runs a non-Exim command in a
6374 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6375 was being done only for the command run by the queryprogram router. It is
6376 now done for all such subprocesses. The other cases are: ${run, transport
6377 filters, and the commands run by the lmtp and pipe transports.
6379 5. Some older OS have a limit of 256 on the maximum number of file
6380 descriptors. Exim was using setrlimit() to set 1000 as a large value
6381 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6382 systems. I've change it so that if it can't get 1000, it tries for 256.
6384 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6385 was an oversight, and furthermore, ever since the addition of extra
6386 controls (e.g. 4.43/32), the checks on when to allow different forms of
6387 "control" were broken. There should now be diagnostics for all cases when a
6388 control that does not make sense is encountered.
6390 7. $recipients is now available in the predata ACL (oversight).
6392 8. Tidy the search cache before the fork to do a delivery from a message
6393 received from the command line. Otherwise the child will trigger a lookup
6394 failure and thereby defer the delivery if it tries to use (for example) a
6395 cached ldap connection that the parent has called unbind on.
6397 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6398 of $address_data from the recipient verification was clobbered by the
6399 sender verification.
6401 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6402 was its contents. (It was OK if the option was not defined at all.)
6404 11. A "Completed" log line is now written for messages that are removed from
6405 the spool by the -Mrm option.
6407 12. $host_address is now set to the target address during the checking of
6408 ignore_target_hosts.
6410 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6411 being passed; this would have caused $sender_host_name to have been used if
6412 matching the list had actually called for a host name (not very likely,
6413 since this list is usually IP addresses). A host name is now passed as
6416 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6417 code that specifies a non-restarting handler (typically sigaction() in
6418 modern systems) in an attempt to fix a rare and obscure crash bug.
6420 15. Narrowed the window for a race in the daemon that could cause it to ignore
6421 SIGCHLD signals. This is not a major problem, because they are used only to
6422 wake it up if nothing else does.
6424 16. A malformed maildirsize file could cause Exim to calculate negative values
6425 for the mailbox size or file count. Odd effects could occur as a result.
6426 The maildirsize information is now recalculated if the size or filecount
6429 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6430 support for a long time. Removed HAVE_SYS_VFS_H.
6432 18. Updated exipick to current release from John Jetmore.
6434 19. Allow an empty sender to be matched against a lookup in an address list.
6435 Previously the only cases considered were a regular expression, or an
6438 20. Exim went into a mad DNS lookup loop when doing a callout where the
6439 host was specified on the transport, if the DNS lookup yielded more than
6442 21. The RFC2047 encoding function was originally intended for short strings
6443 such as real names; it was not keeping to the 75-character limit for
6444 encoded words that the RFC imposes. It now respects the limit, and
6445 generates multiple encoded words if necessary. To be on the safe side, I
6446 have increased the buffer size for the ${rfc2047: expansion operator from
6449 22. Failure to deliver a bounce message always caused it to be frozen, even if
6450 there was an errors_to setting on the router. The errors_to setting is now
6453 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6454 canonical form (fully expanded) before being placed in
6455 $sender_host_address.
6457 24. Updated eximstats to version 1.33
6459 25. Include certificate and key file names in error message when GnuTLS fails
6460 to set them up, because the GnuTLS error message doesn't include the name
6461 of the failing file when there is a problem reading it.
6463 26. Expand error message when OpenSSL has problems setting up cert/key files.
6466 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6467 (this can affect the format of dates).
6469 28. exim_tidydb, when checking for the continued existence of a message for
6470 which it has found a message-specific retry record, was not finding
6471 messages that were in split spool directories. Consequently, it was
6472 deleting retry records that should have stayed in existence.
6474 29. eximstats updated to version 1.35
6475 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6476 1.35 - bugfix such that pie charts by volume are generated correctly
6478 30. The SPA authentication driver was not abandoning authentication and moving
6479 on to the next authenticator when an expansion was forced to fail,
6480 contradicting the general specification for all authenticators. Instead it
6481 was generating a temporary error. It now behaves as specified.
6483 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6484 (the order specifies the preference for clients). The order is now AES256,
6485 AES128, 3DES, ARCFOUR128.
6487 31. Small patch to Sieve code - explicitly set From: when generating an
6490 32. Exim crashed if a remote delivery caused a very long error message to be
6491 recorded - for instance if somebody sent an entire SpamAssassin report back
6492 as a large number of 550 error lines. This bug was coincidentally fixed by
6493 increasing the size of one of Exim's internal buffers (big_buffer) that
6494 happened as part of the Exiscan merge. However, to be on the safe side, I
6495 have made the code more robust (and fixed the comments that describe what
6498 33. Some experimental protocols are using DNS PTR records for new purposes. The
6499 keys for these records are domain names, not reversed IP addresses. The
6500 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6501 leaves it alone. Component reversal etc. now happens only for IP addresses.
6504 34. The host_aton() function is supposed to be passed a string that is known
6505 to be a valid IP address. However, in the case of IPv6 addresses, it was
6506 not checking this. This is a hostage to fortune. Exim now panics and dies
6507 if the condition is not met. A case was found where this could be provoked
6508 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6509 components; fortuitously, this particular loophole had already been fixed
6510 by change 4.50/55 or 4.44/33 above.
6512 If there are any other similar loopholes, the new check in host_aton()
6513 itself should stop them being exploited. The report I received stated that
6514 data on the command line could provoke the exploit when Exim was running as
6515 exim, but did not say which command line option was involved. All I could
6516 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6517 running as the user.
6520 35. There was a buffer overflow vulnerability in the SPA authentication code
6521 (which came originally from the Samba project). I have added a test to the
6522 spa_base64_to_bits() function which I hope fixes it.
6525 36. The daemon start-up calls getloadavg() while still root for those OS that
6526 need the first call to be done as root, but it missed one case: when
6527 deliver_queue_load_max is set with deliver_drop_privilege. This is
6528 necessary for the benefit of the queue runner, because there is no re-exec
6529 when deliver_drop_privilege is set.
6531 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6532 was in use, was not putting the data itself into the right store pool;
6533 consequently, it could be overwritten for a subsequent message in the same
6534 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6537 38. Sometimes the final signoff response after QUIT could fail to get
6538 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6539 < 0 before doing a fflush(). This bug looks as though it goes back to the
6540 introduction of TLS in release 3.20, but "sometimes" must have been rare
6541 because the tests only now provoked it.
6547 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6548 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6549 Now it returns an integer. A number of calls were still expecting a T/F
6550 return. Fortuitously, in all cases, the tests worked in OK situations,
6551 which is the norm. However, things would have gone wrong on any write
6552 failures on the smtp file descriptor. This function is used when sending
6553 messages over SMTP and also when doing verify callouts.
6555 2. When Exim is called to do synchronous delivery of a locally submitted
6556 message (the -odf or -odi options), it no longer closes stderr before doing
6559 3. Implemented the mua_wrapper option.
6561 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6563 5. Implemented the functions header_remove(), header_testname(),
6564 header_add_at_position(), and receive_remove_recipient(), and exported them
6567 6. If an ACL "warn" statement specified the addition of headers, Exim already
6568 inserted X-ACL-Warn: at the start if there was no header name. However, it
6569 was not making this test for the second and subsequent header lines if
6570 there were newlines in the string. This meant that an invalid header could
6571 be inserted if Exim was badly configured.
6573 7. Allow an ACL "warn" statement to add header lines at the start or after all
6574 the Received: headers, as well as at the end.
6576 8. Added the rcpt_4xx retry error code.
6578 9. Added postmaster_mailfrom=xxx to callout verification option.
6580 10. Added mailfrom=xxxx to the callout verification option, for verify=
6583 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6584 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6586 12. Inserted some casts to stop certain compilers warning when using pointer
6587 differences as field lengths or precisions in printf-type calls (mostly
6588 affecting debugging statements).
6590 13. Added optional readline() support for -be (dynamically loaded).
6592 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6593 same clock tick as a message's arrival, so that its received time was the
6594 same as the "first fail" time on the retry record, and that message
6595 remained on the queue past the ultimate address timeout, every queue runner
6596 would try a delivery (because it was past the ultimate address timeout) but
6597 after another failure, the ultimate address timeout, which should have then
6598 bounced the address, did not kick in. This was a "< instead of <=" error;
6599 in most cases the first failure would have been in the next clock tick
6600 after the received time, and all would be well.
6602 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6603 being recognized when the domain list was tested by the match_domain
6604 condition in an expansion string.
6606 16. Added the ${str2b64: operator.
6608 17. Exim was always calling setrlimit() to set a large limit for the number of
6609 processes, without checking whether the existing limit was already
6610 adequate. (It did check for the limit on file descriptors.) Furthermore,
6611 errors from getrlimit() and setrlimit() were being ignored. Now they are
6612 logged to the main and panic logs, but Exim does carry on, to try to do its
6613 job under whatever limits there are.
6615 18. Imported PCRE 5.0.
6617 19. Trivial typo in log message " temporarily refused connection" (the leading
6620 20. If the log selector return_path_on_delivery was set and an address was
6621 redirected to /dev/null, the delivery process crashed because it assumed
6622 that a return path would always be set for a "successful" delivery. In this
6623 case, the whole delivery is bypassed as an optimization, and therefore no
6626 21. Internal re-arrangement: the function for sending a challenge and reading
6627 a response while authentication was assuming a zero-terminated challenge
6628 string. It's now changed to take a pointer and a length, to allow for
6629 binary data in such strings.
6631 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6633 23. Exim was not respecting finduser_retries when seeking the login of the
6634 uid under which it was called; it was always trying 10 times. (The default
6635 setting of finduser_retries is zero.) Also, it was sleeping after the final
6636 failure, which is pointless.
6638 24. Implemented tls_on_connect_ports.
6640 25. Implemented acl_smtp_predata.
6642 26. If the domain in control=submission is set empty, Exim assumes that the
6643 authenticated id is a complete email address when it generates From: or
6644 Sender: header lines.
6646 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6647 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6648 chown and chgrp in /bin and hostname in /usr/bin.
6650 28. Exim was keeping the "process log" file open after each use, just as it
6651 does for the main log. This opens the possibility of it remaining open for
6652 long periods when the USR1 signal hits a daemon. Occasional processlog
6653 errors were reported, that could have been caused by this. Anyway, it seems
6654 much more sensible not to leave this file open at all, so that is what now
6657 29. The long-running daemon process does not normally write to the log once it
6658 has entered its main loop, and it closes the log before doing so. This is
6659 so that log files can straightforwardly be renamed and moved. However,
6660 there are a couple of unusual error situations where the daemon does write
6661 log entries, and I had neglected to close the log afterwards.
6663 30. The text of an SMTP error response that was received during a remote
6664 delivery was being truncated at 512 bytes. This is too short for some of
6665 the long messages that one sometimes sees. I've increased the limit to
6668 31. It is now possible to make retry rules that apply only when a message has a
6669 specific sender, in particular, an empty sender.
6671 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6672 it possible to be selective about when SMTP synchronization is enforced.
6674 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6676 32. Implemented hosts_connection_nolog.
6678 33. Added an ACL for QUIT.
6680 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6683 35. Added mailbox_size and mailbox_filecount to appendfile.
6685 36. Added control = no_multiline_responses to ACLs.
6687 37. There was a bug in the logic of the code that waits for the clock to tick
6688 in the case where the clock went backwards by a substantial amount such
6689 that the microsecond fraction of "now" was more than the microsecond
6690 fraction of "then" (but the whole seconds number was less).
6692 38. Added support for the libradius Radius client library this is found on
6693 FreeBSD (previously only the radiusclient library was supported).
6699 1. When certain lookups returned multiple values in the form name=value, the
6700 quoting of the values was not always being done properly. Specifically:
6701 (a) If the value started with a double quote, but contained no whitespace,
6703 (b) If the value contained whitespace other than a space character (i.e.
6704 tabs or newlines or carriage returns) it was not quoted.
6705 This fix has been applied to the mysql and pgsql lookups by writing a
6706 separate quoting function and calling it from the lookup code. The fix
6707 should probably also be applied to nisplus, ibase and oracle lookups, but
6708 since I cannot test any of those, I have not disturbed their existing code.
6710 2. A hit in the callout cache for a specific address caused a log line with no
6711 reason for rejecting RCPT. Now it says "Previous (cached) callout
6712 verification failure".
6714 3. There was an off-by-one bug in the queryprogram router. An over-long
6715 return line was truncated at 256 instead of 255 characters, thereby
6716 overflowing its buffer with the terminating zero. As well as fixing this, I
6717 have increased the buffer size to 1024 (and made a note to document this).
6719 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6720 when Exim is waiting for an SMTP response from a remote server, Exim
6721 restarts its select() call on the socket, thereby resetting its timeout.
6722 This is not a problem when such interrupts are rare. Somebody set up a cron
6723 job to run exiwhat every 2 minutes, which is less than the normal select()
6724 timeout (5 or 10 minutes). This meant that the select() timeout never
6725 kicked in because it was always reset. I have fixed this by comparing the
6726 time when an interrupt arrives with the time at the start of the first call
6727 to select(). If more time than the timeout has elapsed, the interrupt is
6728 treated as a timeout.
6730 5. Some internal re-factoring in preparation for the addition of Sieve
6731 extensions (by MH). In particular, the "personal" test is moved to a
6732 separate function, and given an option for scanning Cc: and Bcc: (which is
6733 not set for Exim filters).
6735 6. When Exim created an email address using the login of the caller as the
6736 local part (e.g. when creating a From: or Sender: header line), it was not
6737 quoting the local part when it contained special characters such as @.
6739 7. Installed new OpenBSD configuration files.
6741 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6742 try to make them clearer.
6744 9. Callout options, other than the timeout value, were being ignored when
6745 verifying sender addresses in header lines. For example, when using
6747 verify = header_sender/callout=no_cache
6749 the cache was (incorrectly) being used.
6751 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6752 only the Cygwin environment.
6754 11. When return_path_on_delivery was set as a log selector, if different remote
6755 addresses in the same message used different return paths and parallel
6756 remote delivery occurred, the wrong values would sometimes be logged.
6757 (Whenever a remote delivery process finished, the return path value from
6758 the most recently started remote delivery process was logged.)
6760 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6761 lines when AUTH and/or TLS are in use. This is the "received protocol"
6762 field. Exim used to use "asmtp" for authenticated SMTP, without any
6763 indication (in the protocol name) for TLS use. Now it follows the RFC and
6764 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6765 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6766 names appear in log lines as well as in Received: header lines.
6768 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6769 extensions, and comparison tests, and to fix some bugs.
6771 14. Changes to the "personal" filter test:
6773 (1) The test was buggy in that it was just doing the equivalent of
6774 "contains" tests on header lines. For example, if a user's address was
6775 anne@some.where, the "personal" test would incorrectly be true for
6777 To: susanne@some.where
6779 This test is now done by extracting each address from the header in turn,
6780 and checking the entire address. Other tests that are part of "personal"
6781 are now done using regular expressions (for example, to check local parts
6782 of addresses in From: header lines).
6784 (2) The list of non-personal local parts in From: addresses has been
6785 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6786 taken from the Sieve specification recommendations.
6788 (3) If the message contains any header line starting with "List-" it is
6789 treated as non-personal.
6791 (4) The test for "circular" in the Subject: header line has been removed
6792 because it now seems ill-conceived.
6794 15. Minor typos in src/EDITME comments corrected.
6796 16. Installed latest exipick from John Jetmore.
6798 17. If headers_add on a router specified a text string that was too long for
6799 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6800 of string_sprintf() is now avoided.
6802 18. $message_body_size was not set (it was always zero) when running the DATA
6803 ACL and the local_scan() function.
6805 19. For the "mail" command in an Exim filter, no default was being set for
6806 the once_repeat time, causing a random time value to be used if "once" was
6807 specified. (If the value happened to be <= 0, no repeat happened.) The
6808 default is now 0s, meaning "never repeat". The "vacation" command was OK
6809 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6810 (I found it when inspecting the code).
6812 20. There is now an overall timeout for performing a callout verification. It
6813 defaults to 4 times the callout timeout, which applies to individual SMTP
6814 commands during the callout. The overall timeout applies when there is more
6815 than one host that can be tried. The timeout is checked before trying the
6816 next host. This prevents very long delays if there are a large number of
6817 hosts and all are timing out (e.g. when the network connections are timing
6818 out). The value of the overall timeout can be changed by specifying an
6819 additional sub-option for "callout", called "maxwait". For example:
6821 verify = sender/callout=5s,maxwait=20s
6823 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6824 to the end before writing, but this should make it even safer).
6826 22. Exim was forgetting that it had advertised PIPELINING for the second and
6827 subsequent messages on an SMTP connection. It was also not resetting its
6828 memory on STARTTLS and an internal HELO.
6830 23. When Exim logs an SMTP synchronization error within a session, it now
6831 records whether PIPELINING has been advertised or not.
6833 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6834 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6835 rather than long int.
6837 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6839 26. Added the never_mail option to autoreply.
6845 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6846 crash if the getsockname() call failed; this can happen if a connection is
6847 closed very soon after it is established. The problem was simply in the
6848 order in which certain operations were done, causing Exim to try to write
6849 to the SMTP stream before it had set up the file descriptor. The bug has
6850 been fixed by making things happen in the correct order.
6856 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6857 before the connection was closed, thus losing the rejection response.
6859 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6860 some early Solaris releases, but causes trouble in current releases where
6861 socklen_t is defined.
6863 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6866 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6869 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6870 that releasing the top of it at the end releases what was used for sub-
6871 expansions (unless the block got too big). However, discard this block if
6872 the first thing is a variable or header, so that we can use its block when
6873 it is dynamic (useful for very large $message_headers, for example).
6875 6. Lookups now cache *every* query, not just the most recent. A new, separate
6876 store pool is used for this. It can be recovered when all lookup caches are
6877 flushed. Lookups now release memory at the end of their result strings.
6878 This has involved some general refactoring of the lookup sources.
6880 7. Some code has been added to the store_xxx() functions to reduce the amount
6881 of flapping under certain conditions.
6883 8. log_incoming_interface used to affect only the <= reception log lines. Now
6884 it causes the local interface and port to be added to several more SMTP log
6885 lines, for example "SMTP connection from", and rejection lines.
6887 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6889 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6891 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6892 could be overwritten at the end of the current message (or the start of a
6893 new message if it was set in a HELO ACL). The value is now preserved for
6894 the duration of the SMTP connection.
6896 12. If a transport had a headers_rewrite setting, and a matching header line
6897 contained an unqualified address, that address was qualified, even if it
6898 did not match any rewriting rules. The underlying bug was that the values
6899 of the flags that permit the existence of unqualified sender and recipient
6900 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6901 non-local messages, and by -bnq for local messages) were not being
6902 preserved with the message after it was received.
6904 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6905 "next input=" as part of the text comprising the host identity instead of
6906 the correct text. The code was using the same buffer for two different
6907 strings. However, depending on which order the printing function evaluated
6908 its arguments, the bug did not always show up. Under Linux, for example, my
6909 test suite worked just fine.
6911 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6912 doesn't work with some older versions of Perl. It has been changed to "my",
6913 which in any case is probably the better facility to use.
6915 15. A really picky compiler found some instances of statements for creating
6916 error messages that either had too many or two few arguments for the format
6919 16. The size of the buffer for calls to the DNS resolver has been increased
6920 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6921 for addresses that have a lot of PTR records. This alleviates a problem; it
6922 does not fully solve it.
6924 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6925 buffer now truncates the list and logs the incident, which is the same
6926 action as happens when Exim is looking up a host name and its aliases.
6927 Previously in this situation something unpredictable would happen;
6928 sometimes it was "internal error: store_reset failed".
6930 18. If a server dropped the connection unexpectedly when an Exim client was
6931 using GnuTLS and trying to read a response, the client delivery process
6932 crashed while trying to generate an error log message.
6934 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6935 string, for example:
6937 warn message = H1: something\nH2: something
6939 the text was added as a single header line from Exim's point of view
6940 though it ended up OK in the delivered message. However, searching for the
6941 second and subsequent header lines using $h_h2: did not work. This has been
6942 fixed. Similarly, if a system filter added multiple headers in this way,
6943 the routers could not see them.
6945 20. Expanded the error message when iplsearch is called with an invalid key to
6946 suggest using net-iplsearch in a host list.
6948 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6949 ACLs are no longer actually imposed (and a message to that effect is
6952 22. If a "gecos" field in a passwd entry contained escaped characters, in
6953 particular, if it contained a \" sequence, Exim got it wrong when building
6954 a From: or a Sender: header from that name. A second bug also caused
6955 incorrect handling when an unquoted " was present following a character
6956 that needed quoting.
6958 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6959 was not being matched caselessly.
6961 24. Arranged for all hyphens in the exim.8 source to be escaped with
6964 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6965 properly. Recipient callout cache records were still being keyed to include
6966 the sender, even when use_sender was set false. This led to far more
6967 callouts that were necessary. The sender is no longer included in the key
6968 when use_sender is false.
6970 26. Added "control = submission" modifier to ACLs.
6972 27. Added the ${base62d: operator to decode base 62 numbers.
6974 28. dnsdb lookups can now access SRV records.
6976 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6977 the configuration file.
6979 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6980 (-v) mode. This makes the output for a verbose queue run more intelligible.
6982 31. Added a use_postmaster feature to recipient callouts.
6984 32. Added the $body_zerocount variable, containing the number of binary zero
6985 bytes in the message body.
6987 33. The time of last modification of the "new" subdirectory is now used as the
6988 "mailbox time last read" when there is a quota error for a maildir
6991 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6993 35. Added +ignore_unknown as a special item in host lists.
6995 36. Code for decoding IPv6 addresses in host lists is now included, even if
6996 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6997 address was recognized as an IP address, but was then not correctly decoded
6998 into binary, causing unexpected and incorrect effects when compared with
7005 1. Very minor rewording of debugging text in manualroute to say "list of
7006 hosts" instead of "hostlist".
7008 2. If verify=header_syntax was set, and a header line with an unqualified
7009 address (no domain) and a large number of spaces between the end of the
7010 name and the colon was received, the reception process suffered a buffer
7011 overflow, and (when I tested it) crashed. This was caused by some obsolete
7012 code that should have been removed. The fix is to remove it!
7014 3. When running in the test harness, delay a bit after writing a bounce
7015 message to get a bit more predictability in the log output.
7017 4. Added a call to search_tidyup() just before forking a reception process. In
7018 theory, someone could use a lookup in the expansion of smtp_accept_max_
7019 per_host which, without the tidyup, could leave open a database connection.
7021 5. Added the variables $recipient_data and $sender_data which get set from a
7022 lookup success in an ACL "recipients" or "senders" condition, or a router
7023 "senders" option, similar to $domain_data and $local_part_data.
7025 6. Moved the writing of debug_print from before to after the "senders" test
7028 7. Change 4.31/66 (moving the time when the Received: is generated) caused
7029 problems for message scanning, either using a data ACL, or using
7030 local_scan() because the Received: header was not generated till after they
7031 were called (in order to set the time as the time of reception completion).
7032 I have revised the way this works. The header is now generated after the
7033 body is received, but before the ACL or local_scan() are called. After they
7034 are run, the timestamp in the header is updated.
7040 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
7041 before starting a queue runner without re-exec. This happened only when
7042 deliver_drop_privilege was set or when the Exim user was set to root. The
7043 effect of the bug was that timeouts during subsequent deliveries caused
7044 crashes instead of being properly handled. The handler is now left at its
7045 default (and expected) setting.
7047 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
7048 message, again when deliver_drop_privilege is set or Exim is run as root.
7049 The bug described in (1) was not present in this case, but the tidying up
7050 of the other signals was missing. I have made the two cases consistent.
7052 3. The ignore_target_hosts setting on a manualroute router was being ignored
7053 for hosts that were looked up using the /MX notation.
7055 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
7058 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
7059 operated on the sender address. After changing the $sender_address to <>
7060 for the sender address verify, Exim was re-instated it as the original
7061 (before rewriting) address, but remembering that it had rewritten it, so it
7062 wasn't rewriting it again. This bug also had the effect of breaking the
7063 sender address verification caching when the sender address was rewritten.
7065 6. The ignore_target_hosts option was being ignored by the ipliteral router.
7066 This has been changed so that if the ip literal address matches
7067 ignore_target_hosts, the router declines.
7069 7. Added expansion conditions match_domain, match_address, and match_local_
7070 part (NOT match_host).
7072 8. The placeholder for the Received: header didn't have a length field set.
7074 9. Added code to Exim itself and to exim_lock to test for a specific race
7075 condition that could lead to file corruption when using MBX delivery. The
7076 issue is with the lockfile that is created in /tmp. If this file is removed
7077 after a process has opened it but before that process has acquired a lock,
7078 there is the potential for a second process to recreate the file and also
7079 acquire a lock. This could lead to two Exim processes writing to the file
7080 at the same time. The added code performs the same test as UW imapd; it
7081 checks after acquiring the lock that its file descriptor still refers to
7082 the same named file.
7084 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7085 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7088 11. Added the smtp_active_hostname option. If used, this will typically be made
7089 to depend on the incoming interface address. Because $interface_address is
7090 not set up until the daemon has forked a reception process, error responses
7091 that can happen earlier (such as "too many connections") no longer contain
7094 12. If an expansion in a condition on a "warn" statement fails because a lookup
7095 defers, the "warn" statement is abandoned, and the next ACL statement is
7096 processed. Previously this caused the whole ACL to be aborted.
7098 13. Added the iplsearch lookup type.
7100 14. Added ident_timeout as a log selector.
7102 15. Added tls_certificate_verified as a log selector.
7104 16. Added a global option tls_require_ciphers (compare the smtp transport
7105 option of the same name). This controls incoming TLS connections.
7107 17. I finally figured out how to make tls_require_ciphers do a similar thing
7108 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7109 before starting the TLS session.
7111 18. Tabs are now shown as \t in -bP output.
7113 19. If the log selector return_path_on_delivery was set, Exim crashed when
7114 bouncing a message because it had too many Received: header lines.
7116 20. If two routers both had headers_remove settings, and the first one included
7117 a superfluous trailing colon, the final name in the first list and the
7118 first name in the second list were incorrectly joined into one item (with a
7119 colon in the middle).
7125 1. Added -C and -D options to the exinext utility, mainly to make it easier
7126 to include in the automated testing, but these could be helpful when
7127 multiple configurations are in use.
7129 2. The exinext utility was not formatting the output nicely when there was
7130 an alternate port involved in the retry record key, nor when there was a
7131 message id as well (for retries that were specific to a specific message
7132 and a specific host). It was also confused by IPv6 addresses, because of
7133 the additional colons they contain. I have fixed the IPv4 problem, and
7134 patched it up to do a reasonable job for IPv6.
7136 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7137 delivery, the log line now contains "pipelined" if PIPELINING was used.
7139 4. An SMTP transport process used to panic and die if the bind() call to set
7140 an explicit outgoing interface failed. This has been changed; it is now
7141 treated in the same way as a connect() failure.
7143 5. A reference to $sender_host_name in the part of a conditional expansion
7144 that was being skipped was still causing a DNS lookup. This no longer
7147 6. The def: expansion condition was not recognizing references to header lines
7148 that used bh_ and bheader_.
7150 7. Added the _cache feature to named lists.
7152 8. The code for checking quota_filecount in the appendfile transport was
7153 allowing one more file than it should have been.
7155 9. For compatibility with Sendmail, the command line option
7163 and sets the incoming protocol and host name (for trusted callers). The
7164 host name and its colon can be omitted when only the protocol is to be set.
7165 Note the Exim already has two private options, -pd and -ps, that refer to
7166 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7167 "s", but I don't think that's a major issue.
7169 10. A number of refactoring changes to the code, none of which should affect
7172 (a) The number of logging options was getting close to filling up the
7173 32-bit word that was used as a bit map. I have split them into two classes:
7174 those that are passed in the argument to log_write(), and those that are
7175 only ever tested independently outside of that function. These are now in
7176 separate 32-bit words, so there is plenty of room for expansion again.
7177 There is no change in the user interface or the logging behaviour.
7179 (b) When building, for example, log lines, the code previously used a
7180 macro that called string_cat() twice, in order to add two strings. This is
7181 not really sufficiently general. Furthermore, there was one instance where
7182 it was actually wrong because one of the argument was used twice, and in
7183 one call a function was used. (As it happened, calling the function twice
7184 did not affect the overall behaviour.) The macro has been replaced by a
7185 function that can join an arbitrary number of extra strings onto a growing
7188 (c) The code for expansion conditions now uses a table and a binary chop
7189 instead of a serial search (which was left over from when there were very
7190 few conditions). Also, it now recognizes conditions like "pam" even when
7191 the relevant support is not compiled in: a suitably worded error message is
7192 given if an attempt is made to use such a condition.
7194 11. Added ${time_interval:xxxxx}.
7196 12. A bug was causing one of the ddress fields not to be passed back correctly
7197 from remote delivery subprocesses. The field in question was not being
7198 subsequently used, so this caused to problems in practice.
7200 13. Added new log selectors queue_time and deliver_time.
7202 14. Might have fixed a bug in maildirsizefile handling that threw up
7203 "unexpected character" debug warnings, and recalculated the data
7204 unnecessarily. In any case, I expanded the warning message to give more
7207 15. Added the message "Restricted characters in address" to the statements in
7208 the default ACL that block characters like @ and % in local parts.
7210 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7211 Three changes have been made:
7213 (a) There was a serious bug; a negative response to MAIL caused the whole
7214 recipient domain to be cached as invalid, thereby blocking all messages
7215 to all local parts at the same domain, from all senders. This bug has
7216 been fixed. The domain is no longer cached after a negative response to
7217 MAIL if the sender used is not empty.
7219 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7222 (c) A new callout option, "use_sender" has been added for people who want
7223 the modified behaviour.
7229 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7232 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7233 indeed breaks things for older releases.
7235 3. Added additional logging to the case where there is a problem reading data
7236 from a filter that is running in a subprocess using a pipe, in order to
7237 try to track down a specific problem.
7239 4. Testing facility fudge: when running in the test harness and attempting
7240 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7241 getting "No route to host". Convert this to a timeout.
7243 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7246 6. Some OS don't have socklen_t but use size_t instead. This affects the
7247 fifth argument of getsockopt() amongst other things. This is now
7248 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7249 can be set for individual OS. I have set it for SunOS5, OSF1, and
7250 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7251 some earlier ones do not.
7253 7. Change 4.30/15 was not doing the test caselessly.
7255 8. The standard form for an IPv6 address literal was being rejected by address
7256 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7257 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7258 this, as well as the form without the "IPv6" on the front (but only when
7259 address literals are enabled, of course).
7261 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7263 10. Exim crashed if a message with an empty sender address specified by -f
7264 encountered a router with an errors_to setting. This could be provoked only
7265 by a command such as
7269 where an empty string was supplied; "<>" did not hit this bug.
7271 11. Installed PCRE release 4.5.
7273 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7274 remained set. It is now erased.
7276 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7277 times from message ids (which are base 36 rather than the normal 62).
7279 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7280 were being counted as actual protocol errors, and logged if the log
7281 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7282 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7283 and DATA following a set of rejected RCPTs do not count as protocol errors.
7284 In other words, Exim assumes they were pipelined, though this may not
7285 actually be the case. Of course, in all cases the client gets an
7286 appropriate error code.
7288 15. If a lookup fails in an ACL condition, a message about the failure may
7289 be available; it is used if testing the ACL cannot continue, because most
7290 such messages specify what the cause of the deferral is. However, some
7291 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7292 that caused an old message to be retained and used if a later statement
7293 caused a defer, replacing the real cause of the deferral.
7295 16. If an IP address had so many PTR records that the DNS lookup buffer
7296 was not large enough to hold them, Exim could crash while trying to process
7297 the truncated data. It now detects and logs this case.
7299 17. Further to 4.21/58, another change has been made: if (and only if) the
7300 first line of a message (the first header line) ends with CRLF, a bare LF
7301 in a subsequent header line has a space inserted after it, so as not to
7302 terminate the header.
7304 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7305 unnecessarily, used atoi() instead of strtol(), and didn't check the
7306 termination when getting file sizes from file names by regex.
7308 19. Completely re-implemented the support for maildirsize files, in the light
7309 of a number of problems with the previous contributed implementation
7310 (4.30/29). In particular:
7312 . If the quota is zero, the maildirsize file is maintained, but no quota is
7315 . If the maildir directory does not exist, it is created before any attempt
7316 to write a maildirsize file.
7318 . The quota value in the file is just a cache; if the quota is changed in
7319 the transport, the new value overrides.
7321 . A regular expression is available for excluding directories from the
7324 20. The autoreply transport checks the characters in options that define the
7325 message's headers; it allows continued headers, but it was checking with
7326 isspace() after an embedded newline instead of explicitly looking for a
7329 21. If all the "regular" hosts to which an address was routed had passed their
7330 expiry times, and had not reached their retry times, the address was
7331 bounced, even if fallback hosts were defined. Now Exim should go on to try
7334 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7335 equivalent code in the SMTP transport. Some hosts send humungous responses
7336 to HELO/EHLO, more than 1024 it seems.
7338 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7339 gives compiler warnings in some environments. I've now done it "properly",
7342 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7343 (because of problems with the built-in one) was declared to return uschar *
7344 instead of char *, causing compiler failure.
7346 25. Fixed a file descriptor leak when processing alias/forward files.
7348 26. Fixed a minor format string issue in dbfn.c.
7350 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7352 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7353 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7356 29. When a sender address is verified, it is cached, to save repeating the test
7357 when there is more than one recipient in a message. However, when the
7358 verification involves a callout, it is possible for different callout
7359 options to be set for different recipients. It is too complicated to keep
7360 track of this in the cache, so now Exim always runs a verification when a
7361 callout is required, relying on the callout cache for the optimization.
7362 The overhead is duplication of the address routing, but this should not be
7365 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7366 to be verified with callout=postmaster, and the main callout worked but the
7367 postmaster check failed, the verification correctly failed. However, if a
7368 subsequent RCPT command asked for sender verification *without* the
7369 postmaster check, incorrect caching caused this verification also to fail,
7372 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7373 it was not caching the DNS options (qualify_single, search_parents) that
7374 were used when the lookup failed. A subsequent lookup with different
7375 options therefore always gave the same answer, though there were cases
7376 where it should not have. (Example: a "domains = !$mx_any" option on a
7377 dnslookup router: the "domains" option is always processed without any
7378 widening, but the router might have qualify_single set.) Now Exim uses the
7379 cached value only when the same options are set.
7381 32. Added John Jetmore's "exipick" utility to the distribution.
7383 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7384 than a timeout (e.g. a certificate is required, and is not provided), an
7385 Exim server now closes the connection immediately. Previously it waited for
7386 the client to close - but if the client is SSL, it seems that they each
7387 wait for each other, leading to a delay before one of them times out.
7389 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7390 maintained 0.8.x compatibility because I don't think many are using it, and
7391 it is clearly obsolete.
7393 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7396 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7397 client certificate was expired. A simple patch fixes this, though I don't
7398 understand the full logic of why the verify callback is called multiple
7401 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7402 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7403 which causes problems with some clients (such as the Certicom SSL Plus
7404 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7405 disables the coutermeasure allowing Eudora to connect."
7407 38. Exim was not checking that a write() to a log file succeeded. This could
7408 lead to Bad Things if a log got too big, in particular if it hit a file
7409 size limit. Exim now panics and dies if it cannot write to a log file, just
7410 as it does if it cannot open a log file.
7412 39. Modified OS/Makefile-Linux so that it now contains
7414 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7416 The two -D definitions ensure that Exim is compiled with large file
7417 support, which makes it possible to handle log files that are bigger than
7420 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7421 instance) a domain was checked against a named list that involved a lookup,
7422 causing $domain_data to be set, then another domain was checked against the
7423 same list, then the first domain was re-checked, the value of $domain_data
7424 after the final check could be wrong. In particular, if the second check
7425 failed, it could be set empty. This bug probably also applied to
7428 41. The strip_trailing_dot option was not being applied to the address given
7429 with the -f command-line option.
7431 42. The code for reading a message's header from the spool was incrementing
7432 $received_count, but never initializing it. This meant that the value was
7433 incorrect (doubled) while delivering a message in the same process in which
7434 it was received. In the most common configuration of Exim, this never
7435 happens - a fresh exec is done - but it can happen when
7436 deliver_drop_privilege is set.
7438 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7439 it now includes up to 150 characters of the unexpected data in the log
7442 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7443 and building data strings. The size of both of these buffers was 10 000
7444 bytes - far larger than anybody would *ever* want, thought I. Needless to
7445 say, somebody hit the limit. I have increased the maximum line length to
7446 20 000 and the maximum data length of concatenated lines to 100 000. I have
7447 also fixed two bugs, because there was no checking on these buffers. Tsk,
7448 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7449 buffer is too small.
7451 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7452 lsearch lookups. Now it does.
7454 46. When parsing a route_list item in a manualroute router, a fixed-length
7455 buffer was used for the list of hosts. I made this 1024 bytes long,
7456 thinking that nobody would ever have a list of hosts that long. Wrong.
7457 Somebody had a whole pile of complicated expansion conditions, and the
7458 string was silently truncated, leading to an expansion error. It turns out
7459 that it is easier to change to an unlimited length (owing to other changes
7460 that have happened since this code was originally written) than to build
7461 structure for giving a limitation error. The length of the item that
7462 expands into the list of hosts is now unlimited.
7464 47. The lsearch lookup could not handle data where the length of text line was
7465 more than 4095 characters. Such lines were truncated, leading to shortened
7466 data being returned. It should now handle lines of any length.
7468 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7469 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7472 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7473 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7474 no longer happens in comment lines. A list of replacements is now placed
7475 at the head of all of the source files, except those whose only change is
7476 to replace PERL_COMMAND in the very first #! line.
7478 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7479 messages on the queue, with a bottom-up merge sort, using code contributed
7480 by Michael Haardt. This should make operations like -bp somewhat faster on
7481 large queues. It won't affect queue runners, except when queue_run_in_order
7484 51. Installed eximstats 1.31 in the distribution.
7486 52. Added support for SRV lookups to the dnslookup router.
7488 53. If an ACL referred to $message_body or $message_body_end, the value was not
7489 reset for any messages that followed in the same SMTP session.
7491 54. The store-handling optimization for building very long strings was not
7492 differentiating between the different store pools. I don't think this
7493 actually made any difference in practice, but I've tidied it.
7495 55. While running the routers to verify a sender address, $sender_address
7496 was still set to the sender address. This is wrong, because when routing to
7497 send a bounce to the sender, it would be empty. Therefore, I have changed
7498 it so that, while verifying a sender address, $sender_address is set to <>.
7499 (There is no change to what happens when verifying a recipient address.)
7501 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7502 target A or AAAA records (if not already returned) without resetting the
7503 qualify_single or search_parents options of the DNS resolver. These are
7504 inappropriate in this case because the targets of MX and SRV records must
7505 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7506 target that, when qualified, matched something in the local domain. These
7507 two options are now turned off when doing these lookups.
7509 57. It seems that at least some releases of Reiserfs (which does not have the
7510 concept of a fixed number of inodes) returns zero and not -1 for the
7511 number of available inodes. This interacted badly with check_spool_inodes,
7512 which assumed that -1 was the "no such thing" setting. What I have done is
7513 to check that the total number of inodes is greater than zero before doing
7514 the test of how many are available.
7516 58. When a "warn" ACL statement has a log_message modifier, the message is
7517 remembered, and not repeated. This is to avoid a lot of repetition when a
7518 message has many recipients that cause the same warning to be written.
7519 However, Exim was preserving the list of already written lines for an
7520 entire SMTP session, which doesn't seem right. The memory is now reset if a
7521 new message is started.
7523 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7524 debugging output unless log_rewrite was also set.
7526 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7527 of (int)(handle) when we know that handle contains (void *)(-1).
7529 61. The Exim daemon panic-logs an error return when it closes the incoming
7530 connection. However "connection reset by peer" seems to be common, and
7531 isn't really an error worthy of noting specially, so that particular error
7534 62. When Exim is trying to find all the local interfaces, it used to panic and
7535 die if the ioctl to get the interface flags failed. However, it seems that
7536 on at least one OS (Solaris 9) it is possible to have an interface that is
7537 included in the list of interfaces, but for which you get a failure error
7538 for this call. This happens when the interface is not "plumbed" into a
7539 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7540 failure of the "get flags" call assumes that the interface is down.
7542 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7543 makes life easier for people who are doing arithmetic on fields extracted
7544 from dates, where you often get leading zeros that should not be
7545 interpreted as octal.
7547 64. Added qualify_domain to the redirect router, to override the global
7550 65. If a pathologically long header line contained very many addresses (the
7551 report of this problem mentioned 10 000) and each of them was rewritten,
7552 Exim could use up a very large amount of memory. (It kept on making new
7553 copies of the header line as it rewrote, and never released the old ones.)
7554 At the expense of a bit more processing, the header rewriting function has
7555 been changed so that it no longer eats memory in this way.
7557 66. The generation of the Received: header has been moved from the time that a
7558 message starts to be received, to the time that it finishes. The timestamp
7559 in the Received: header should now be very close to that of the <= log
7560 line. There are two side-effects of this change:
7562 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7563 logged header lines no longer include the local Received: line, because
7564 it has not yet been created. The same applies to a copy of the message
7565 that is returned to a non-SMTP sender when a message is rejected.
7567 (b) When a filter file is tested using -bf, no additional Received: header
7568 is added to the test message. After some thought, I decided that this
7571 This change does not affect the value of $received_for. It is still set
7572 after address rewriting, but before local_scan() is called.
7574 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7576 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7577 gave an unhelpful panic error message, and a defer error. I have managed to
7578 change this behaviour so that it now rejects any supplied certificate,
7579 which seems right, as the list of acceptable certificates is empty.
7581 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7582 gave an unhelpful defer error. I have not managed to make this reject any
7583 supplied certificates, but the error message it gives is "no certificate
7584 supplied", which is not helpful.
7586 70. exigrep's output now also includes lines that are not associated with any
7587 message, but which match the given pattern. Implemented by a patch from
7588 Martin Sluka, which also tidied up the Perl a bit.
7590 71. Recipient callout verification, like sender verification, was using <> in
7591 the MAIL FROM command. This isn't really the right thing, since the actual
7592 sender may affect whether the remote host accepts the recipient or not. I
7593 have changed it to use the actual sender in the callout; this means that
7594 the cache record is now keyed on a recipient/sender pair, not just the
7595 recipient address. There doesn't seem to be a real danger of callout loops,
7596 since a callout by the remote host to check the sender would use <>.
7597 [SEE ABOVE: changed after hitting problems.]
7599 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7600 temporary errors. However, in the case of such a code being given after
7601 the end of a data transmission (i.e. after ".") Exim was failing to write
7602 a retry record for the message. (Yes, there was some broken host that was
7603 actually sending 8xx at this point.)
7605 73. An unknown lookup type in a host list could cause Exim to panic-die when
7606 the list was checked. (An example that provoked this was putting <; in the
7607 middle of a list instead of at the start.) If this happened during a DATA
7608 ACL check, a -D file could be left lying around. This kind of configuration
7609 error no longer causes Exim to die; instead it causes a defer error. The
7610 incident is still logged to the main and panic logs.
7612 74. Buglet left over from Exim 3 conversion. The message "too many messages
7613 in one connection" was written to the rejectlog but not the mainlog, except
7614 when address rewriting (yes!) was being logged.
7616 75. Added write_rejectlog option.
7618 76. When a system filter was run not as root (that is, when system_filter_user
7619 was set), the values of the $n variables were not being returned to the
7620 main process; thus, they were not subsequently available in the $sn
7623 77. Added +return_path_on_delivery log selector.
7625 78. A connection timeout was being treated differently from recipients deferred
7626 when testing hosts_max_try with a message that was older than the host's
7627 retry timeout. (The host should not be counted, thus allowing all hosts to
7628 be tried at least once before bouncing.) This may have been the cause of an
7629 occasionally reported bug whereby a message would remain on the queue
7630 longer than the retry timeout, but would be bounced if a delivery was
7631 forced. I say "may" because I never totally pinned down the problem;
7632 setting up timeout/retry tests is difficult. See also the next item.
7634 79. The ultimate address timeout was not being applied to errors that involved
7635 a combination of host plus message (for example, a timeout on a MAIL
7636 command). When an address resolved to a number of possible hosts, and they
7637 were not all tried for each delivery (e.g. because of hosts_max_try), a
7638 message could remain on the queue longer than the retry timeout.
7640 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7643 81. Fixed an obscure SMTP outgoing bug which required at least the following
7644 conditions: (a) there was another message waiting for the same server;
7645 (b) the server returned 5xx to all RCPT commands in the first message so
7646 that the message was not completed; (c) the server dropped the connection
7647 or gave a negative response to the RSET that Exim sends to abort the
7648 transaction. The observed case was a dropped connection after DATA that had
7649 been sent in pipelining mode. That is, the server had advertised PIPELINING
7650 but was not implementing it correctly. The effect of the bug was incorrect
7651 behaviour, such as trying another host, and this could lead to a crash.
7657 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7658 and daemon.c were passed as pointers to ints; they should have been
7659 pointers to socklen_t variables (which are typically unsigned ints).
7661 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7664 3. Fixed a really odd bug that affected only the testing scheme; patching a
7665 certain fixed string in the binary changed the value of another string that
7666 happened to be identical to the end of the original first string.
7668 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7669 name", it returns that address as the IP address. On some operating
7670 systems (e.g. Solaris), it also passes back the IP address string as the
7671 "host name". However, on others (e.g. Linux), it passes back an empty
7672 string. Exim wasn't checking for this, and was changing the host name to an
7673 empty string, assuming it had been canonicalized.
7675 5. Although rare, it is permitted to have more than one PTR record for a given
7676 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7677 all the names associated with an address, because they do in Solaris.
7678 However, it seems that they do not in Linux for data that comes from the
7679 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7680 I found this out when I moved to a new Linux workstation and tried to run
7681 the Exim test suite.
7683 To get round this problem I have changed the code so that it now does its
7684 own call to the DNS to look up PTR records when searching for a host name.
7685 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7686 addresses that are only in /etc/hosts are still found.
7688 This behaviour is, however, controlled by an option called host_lookup_
7689 order, which defaults to "bydns:byaddr". If people want to use the other
7690 order, or indeed, just use one or the other means of lookup, they can
7691 specify it in this variable.
7693 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7694 some operating systems, this comes back from gethostbyaddr() as an empty
7695 string, and this is what Exim used to test for. However, it seems that in
7696 other systems, "." is yielded. Exim now tests for this case too.
7698 7. The values of check_spool_space and check_log_space are now held internally
7699 as a number of kilobytes instead of an absolute number of bytes. If a
7700 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7701 kilobyte. This means that much larger values can be stored.
7703 8. Exim monitor: an attempt to get the action menu when not actually pointing
7704 at a message produces an empty menu entitled "No message selected". This
7705 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7706 no entries in it ("Shell widget menu has zero width and/or height"). So I
7707 have added a single, blank menu entry in this case.
7709 9. Added ${quote_local_part.
7711 10. MIME decoding is now applied to the contents of Subject: header lines when
7714 11. Now that a reference to $sender_host_address automatically causes a reverse
7715 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7716 host lookup before query-style lookups in lists that might use this
7717 variable. This has therefore been abolished, and the "net-" prefix is no
7718 longer necessary for query-style lookups.
7720 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7721 have been a typo for LFLAGS, so it has been changed.
7723 13. The install script calls Exim with "-C /dev/null" in order to find the
7724 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7725 to be output. However, since Exim outputs its version number before the
7726 error, it didn't break the script. It just looked ugly. I fixed this by
7727 always allowing "-C /dev/null" if the caller is root.
7729 14. Ignore overlarge ACL variable number when reading spool file - insurance
7730 against a later release with more variables having written the file.
7732 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7733 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7734 this, as well as the form without the "IPv6" on the front.
7736 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7737 OS/Makefile-Darwin file.
7739 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7740 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7742 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7743 variable. However, some OS use ints for inodes. I've added cast to long int
7744 to get rid of the compiler warning.
7746 19. I had forgotten to lock out "/../" in configuration file names when
7747 ALT_CONFIG_PREFIX was set.
7749 20. Routers used for verification do not need to specify transports. However,
7750 if such a router generated a host list, and callout was configured, Exim
7751 crashed, because it could not find a port number from the (non-existent)
7752 transport. It now assumes port 25 in this circumstance.
7754 21. Added the -t option to exigrep.
7756 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7757 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7758 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7759 them is set, LOOKUP_LSEARCH is forced.
7761 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7763 24. Added sender and host information to the "rejected by local_scan()" log
7764 line; previously there was no indication of these.
7766 25. Added .include_if_exists.
7768 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7769 Linux. It turns out that not all file systems support this. Apparently some
7770 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7771 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7772 not supported on the file descriptor, is now ignored when Exim is trying to
7773 sync a directory. This applies only to Linux.
7775 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7777 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7778 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7781 29. Added support for maildirsize files from supplied patch (modified a bit).
7783 30. The use of :fail: followed by an empty string could lead Exim to respond to
7784 sender verification failures with (e.g.):
7786 550 Verification failed for <xxx>
7787 550 Sender verify failed
7789 where the first response line was missing the '-' that indicates it is not
7790 the final line of the response.
7792 31. The loop for finding the name of the user that called Exim had a hardwired
7793 limit of 10; it now uses the value of finduser_retries, which is used for
7794 all other user lookups.
7796 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7799 33. Exim was neglecting to zero errno before one call of strtol() when
7800 expanding a string and expecting an integer value. On some systems this
7801 resulted in spurious "integer overflow" errors. Also, it was casting the
7802 result into an int without checking.
7804 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7805 did not work. The code looks as if it has *never* worked, though it appears
7806 to have been documented since at least release 1.62. I have made it work.
7808 35. The "timeout_DNS" error in retry rules, also documented since at least
7809 1.62, also never worked. As it isn't clear exactly what this means, and
7810 clearly it isn't a major issue, I have abolished the feature by treating it
7811 as "timeout", and writing a warning to the main and panic logs.
7813 36. The display of retry rules for -brt wasn't always showing the error code
7816 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7817 timeout_connect_A, timeout_connect_MX.
7819 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7820 to the empty sender.
7822 39. The daemon was not analysing the content of -oX till after it had closed
7823 stderr and disconnected from the controlling terminal. This meant that any
7824 syntax errors were only noted on the panic log, and the return code from
7825 the command was 0. By re-arranging the code a little, I've made the
7826 decoding happen first, so such errors now appear on stderr, and the return
7827 code is 1. However, the actual setting up of the sockets still happens in
7828 the disconnected process, so errors there are still only recorded on the
7831 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7832 connections (as happens on some IP stacks) was logged at start up time as
7833 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7834 it from "IPv6 and IPv4", which means that two separate sockets are being
7837 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7838 says whether AF_INET or AF_INET6 was passed as an argument.
7840 42. Exiwhat output was messed up when time zones were included in log
7843 43. Exiwhat now gives more information about the daemon's listening ports,
7844 and whether -tls-on-connect was used.
7846 44. The "port" option of the smtp transport is now expanded.
7848 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7849 silently ignored. Now an error message is written to the main and panic
7852 46. There's a new ACL modifier called "logwrite" which writes to a log file
7853 as soon as it is encountered.
7855 47. Added $local_user_uid and $local_user_gid at routing time.
7857 48. Exim crashed when trying to verify a sender address that was being
7860 49. Exim was recognizing only a space character after ".include". It now also
7861 recognizes a tab character.
7863 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7864 extracting the relevant information from the specification. The man page no
7865 longer contains scrambled data for the -d option, and I've added a section
7866 at the front about calling Exim under different names.
7868 51. Added "extra_headers" argument to the "mail" command in filter files.
7870 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7873 53. Installed eximstats 1.29.
7875 54. Added transport_filter_timeout as a generic transport option.
7877 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7878 Cc: header lines. This was required by RFC 822, but it not required by RFC
7881 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7882 incoming messages that did not have them. Now it does so only if the
7883 message originates locally, that is, if there is no associated remote host
7884 address. When Resent- header lines are present, this applies to the Resent-
7885 lines rather than the non-Resent- lines.
7887 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7888 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7890 58. Messages for configuration errors now include the name of the main
7891 configuration file - useful now that there may be more than one file in a
7892 list (.included file names were always shown).
7894 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7895 for those rare installations that do not start the daemon as root or run it
7896 setuid root. I've cut out the call to initgroups() if the daemon is not
7899 60. The Exim user and group can now be bound into the binary as text strings
7900 that are looked up at the start of Exim's processing.
7902 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7904 62. Added $mailstore_basename variable.
7906 63. Installed patch to sieve.c from Michael Haardt.
7908 64. When Exim failed to open the panic log after failing to open the main log,
7909 the original message it was trying to log was written to stderr and debug
7910 output, but if they were not available (the usual case in production), it
7911 was lost. Now it is written to syslog before the two lines that record the
7912 failures to open the logs.
7914 65. Users' Exim filters run in subprocesses under the user's uid. It is
7915 possible for a "deliver" command or an alias in a "personal" command to
7916 provoke an address rewrite. If logging of address rewriting is configured,
7917 this fails because the process is not running as root or exim. There may be
7918 a better way of dealing with this, but for the moment (because 4.30 needs
7919 to be released), I have disabled address rewrite logging when running a
7920 filter in a non-root, non-exim process.
7926 1. The buildconfig auxiliary program wasn't quoting the value set for
7927 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7928 not defined. This bug was masked in 4.22 by the effect that was fixed in
7931 2. Some messages that were rejected after a message id was allocated were
7932 shown as "incomplete" by exigrep. It no longer does this for messages that
7933 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7935 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7936 have allow_domain_literals set, the ID did not get logged in the <= line.
7937 Domain literals are now always recognized in Message-ID: header lines.
7939 4. The first argument for a ${extract expansion item is the key name or field
7940 number. Leading and trailing spaces in this item were not being ignored,
7941 causing some misleading effects.
7943 5. When deliver_drop_privilege was set, single queue runner processes started
7944 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7945 same command in the process it spins off) were not dropping privilege.
7947 6. When the daemon running as "exim" started a queue runner, it always
7948 re-executed Exim in the spun-off process. This is a waste of effort when
7949 deliver_drop_privilege is set. The new process now just calls the
7950 queue-runner function directly.
7956 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7959 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7960 ignored. Though the use of -oP was forcing the writing of a pid file, it
7961 was always written to the default place.
7963 3. If the message "no IP address found for host xxxx" is generated during
7964 incoming verification, it is now followed by identification of the incoming
7965 connection (so you can more easily find what provoked it).
7967 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7969 5. Added some features to "harden" Exim a bit more against certain attacks:
7971 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7972 be put in Local/Makefile. This is like the never_users runtime option,
7973 but it cannot be overridden. The default setting is "root".
7975 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7976 prefix string with which any file named in a -C command line option
7979 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7980 is retained for -C and -D only if the caller of Exim is root. Without
7981 it, the exim user may also use -C and -D and retain privilege.
7983 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7984 command line option is disabled.
7986 6. Macro names set by the -D option must start with an upper case letter, just
7987 like macro names defined in the configuration file.
7989 7. Added "dereference=" facility to LDAP.
7991 8. Two instances of the typo "uknown" in the source files are fixed.
7993 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7994 the Configure-Makefile script screwed up while processing it.
7996 10. Incorporated PCRE 4.4.
7998 11. The SMTP synchronization check was not operating right at the start of an
7999 SMTP session. For example, it could not catch a HELO sent before the client
8000 waited for the greeting. There is now a check for outstanding input at the
8001 point when the greeting is written. Because of the duplex, asynchronous
8002 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
8003 way, but not yet received, when the check is performed.
8005 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
8006 on TCP/IP sockets, because this apparently causes some broken clients to
8009 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
8010 unchanged) from the Cygwin maintainer.
8012 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
8013 was supported instead of testing for mbx. Effectively a typo.
8015 15. The spa authenticator server code was not checking that the input it
8016 received was valid base64.
8018 16. The debug output line for the "set" modifier in ACLs was not showing the
8019 name of the variable that was being set.
8021 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
8023 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
8024 lookup on its own. Something else was needed to trigger the lookup. For
8025 example, a match in host_lookup or the need for a host name in a host list.
8026 Now, if $sender_host_name is referenced and the host name has not yet been
8027 looked up, a lookup is performed. If the lookup fails, the variable remains
8028 empty, and $host_lookup_failed is set to "1".
8030 19. Added "eqi" as a case-independent comparison operator.
8032 20. The saslauthd authentication condition could segfault if neither service
8033 nor realm was specified.
8035 21. If an overflowing value such as "2048M" was set for message_size_limit, the
8036 error message that was logged was misleading, and incoming SMTP
8037 connections were dropped. The message is now more accurate, and temporary
8038 errors are given to SMTP connections.
8040 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
8041 (except RSET) with a 421 error, until QUIT is received. However, it was
8042 failing to send a response to QUIT.
8044 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
8045 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
8046 ACL is now run after the helo_try_verify_hosts code.
8048 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
8049 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
8050 case-independent, so other case variants are also recognized). Apparently
8051 some systems use these upper case variants.
8053 25. If more than two messages were waiting for the same host, and a transport
8054 filter was specified for the transport, Exim sent two messages over the
8055 same TCP/IP connection, and then failed with "socket operation on non-
8056 socket" when it tried to send the third.
8058 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
8060 27. The extern definition of crypt16() in expand.c was not being excluded when
8061 the OS had its own crypt16() function.
8063 28. Added bounce_return_body as a new option, and bounce_return_size_limit
8064 as a preferred synonym for return_size_limit, both as an option and as an
8067 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8069 30. Changed the default configuration ACL to relax the local part checking rule
8070 for addresses that are not in any local domains. For these addresses,
8071 slashes and pipe symbols are allowed within local parts, but the sequence
8072 /../ is explicitly forbidden.
8074 31. SPA server authentication was not clearing the challenge buffer before
8077 32. log_message in a "warn" ACL statement was writing to the reject log as
8078 well as to the main log, which contradicts the documentation and doesn't
8079 seem right (because no rejection is happening). So I have stopped it.
8081 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8082 However, I am unable to do any testing of this.
8084 34. Fixed an infelicity in the appendfile transport. When checking directories
8085 for a mailbox, to see if any needed to be created, it was accidentally
8086 using path names with one or more superfluous leading slashes; tracing
8087 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8089 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8090 contents are added to the log line that is written for every discarded
8091 recipient. (Previously a log_message setting was ignored.)
8093 36. The ${quote: operator now quotes the string if it is empty.
8095 37. The install script runs exim in order to find its version number. If for
8096 some reason other than non-existence or emptiness, which it checks, it
8097 could not run './exim', it was installing it with an empty version number,
8098 i.e. as "exim-". This error state is now caught, and the installation is
8101 38. An argument was missing from the function that creates an error message
8102 when Exim fails to connect to the socket for saslauthd authentication.
8103 This could cause Exim to crash, or give a corrupted message.
8105 39. Added isip, isip4, and isip6 to ${if conditions.
8107 40. The ACL variables $acl_xx are now saved with the message, and can be
8108 accessed later in routers, transports, and filters.
8110 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8111 strings in the file are not string-expanded.
8113 42. If a MAIL command specified a SIZE value that was too large to fit into an
8114 int variable, the check against message_size_limit failed. Such values are
8115 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8116 day this will have to be increased, but I don't think I want to be around
8117 when emails are that large.
8124 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8125 iconv() is not standard in FreeBSD.
8127 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8128 IPv6 enabled. The observed symptom was a segmentation fault on return from
8129 the function os_common_find_running_interfaces() in src/os.c.
8131 3. In the check_special_case() function in daemon.c I had used "errno" as an
8132 argument name, which causes warnings on some systems. This was basically a
8133 typo, since it was named "eno" in the comments!
8135 4. The code that waits for the clock to tick (at a resolution of some fraction
8136 of a second) so as to ensure message-id uniqueness was always waiting for
8137 at least one whole tick, when it could have waited for less. [This is
8138 almost certainly not relevant at current processor speeds, where it is
8139 unlikely to ever wait at all. But we try to future-proof.]
8141 5. The function that sleeps for a time interval that includes fractions of a
8142 second contained a race. It did not block SIGALRM between setting the
8143 timer, and suspending (a couple of lines later). If the interval was short
8144 and the sigsuspend() was delayed until after it had expired, the suspension
8145 never ended. On busy systems this could lead to processes getting stuck for
8148 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8149 process, before it forks any delivery processes. The open lookup caching
8150 mechanism meant that the open file or database connection was passed into
8151 the delivery process. The problem was that delivery processes always tidy
8152 up cached lookup data. This could cause a problem for the next delivery
8153 process started by the queue runner, because the external queue runner
8154 process does not know about the closure. So the next delivery process
8155 still has data in the lookup cache. In the case of a file lookup, there was
8156 no problem because closing a file descriptor in a subprocess doesn't affect
8157 the parent. However, if the lookup was caching a connection to a database,
8158 the connection was closed, and the second delivery process was likely to
8159 see errors such as "PGSQL: query failed: server closed the connection
8160 unexpectedly". The problem has been fixed by closing all cached lookups
8161 in a queue runner before running a delivery process.
8163 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8164 seem to have the "const" qualifier which it has on other OS. I've
8167 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8168 *of the same type* (client or server) with the same public name that an
8169 error should be diagnosed.
8171 9. When Exim looked up a host name for an IP address, but failed to find the
8172 original IP address when looking up the host name (a safety check), it
8173 output the message "<ip address> does not match any IP for NULL", which was
8174 confusing, to say the least. The bug was that the host name should have
8175 appeared instead of "NULL".
8177 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8178 user that is built into the binary, and the -C or -D options is used, root
8179 privilege is dropped before the configuration file is read. In addition,
8180 logging is switched to stderr instead of the normal log files. If the
8181 configuration then re-defines the Exim user, the unprivileged environment
8182 is probably not what is expected, so Exim logs a panic warning message (but
8185 However, if deliver_drop_privilege is set, the unprivileged state may well
8186 be exactly what is intended, so the warning has been cut out in that case,
8187 and Exim is allowed to try to write to its normal log files.
8193 1. smtp_return_error_details was not giving details for temporary sender
8194 or receiver verification errors.
8196 2. Diagnose a configuration error if two authenticators have the same public
8199 3. Exim used not to create the message log file for a message until the first
8200 delivery attempt. This could be confusing when incoming messages were held
8201 for policy or load reasons. The message log file is now created at the time
8202 the message is received, and an initial "Received" line is written to it.
8204 4. The automatically generated man page for command line options had a minor
8205 bug that caused no ill effects; however, a more serious problem was that
8206 the procedure for building the man page automatically didn't always
8207 operate. Consequently, release 4.20 contains an out-of-date version. This
8208 shouldn't happen again.
8210 5. When building Exim with embedded Perl support, the script that builds the
8211 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8212 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8214 6. The freeze_tell option was not being used for messages that were frozen on
8215 arrival, either by an ACL or by local_scan().
8217 7. Added the smtp_incomplete_transaction log selector.
8219 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8220 was accepting AUTH without a new EHLO.
8222 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8223 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8226 10. Logging of TCP/IP connections (when configured) now happens in the main
8227 daemon process instead of the child process, so that the TCP/IP connection
8228 count is more accurate (but it can never be perfect).
8230 11. The use of "drop" in a nested ACL was not being handled correctly in the
8231 outer ACL. Now, if condition failure induced by the nested "drop" causes
8232 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8233 or "require"), the connection is dropped.
8235 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8236 that yield "discard" can now be used with an "accept" or a "discard" verb,
8237 but an error is generated for any others (because I can't see a useful way
8238 to define what should happen).
8240 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8241 are now processed in the same way as lines in the Exim configuration file.
8242 In particular, continuation lines are supported.
8244 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8246 15. Added -ti meaning -t -i.
8248 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8249 domains, and warn by logging if others are found.
8251 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8252 returned from GIFCONF when Exim is trying to find the list of interfaces on
8253 a host. The code in os.c has been modified to copy each ifreq to an aligned
8254 structure in all cases.
8256 Also, in some cases, the returned ifreq's were being copied to a 'struct
8257 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8258 means the last couple of bytes of an IPv6 address could be chopped if the
8259 ifreq contained only a normal sockaddr (14 bytes storage).
8261 18. Named domain lists were not supported in the hosts_treat_as_local option.
8262 An entry such as +xxxx was not recognized, and was treated as a literal
8265 19. Ensure that header lines added by a DATA ACL are included in the reject log
8266 if the ACL subsequently rejects the message.
8268 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8269 MD5 (which is deprecated).
8271 21. When testing a filter file using -bf, Exim was writing a message when it
8272 took the sender from a "From " line in the message, but it was not doing so
8273 when it took $return_path from a Return-Path: header line. It now does.
8275 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8276 with a valid header line field name (a series of printing characters
8277 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8279 23. Changed "disc" in the source to "disk" to conform to the documentation and
8280 the book and for uniformity.
8282 24. Ignore Sendmail's -Ooption=value command line item.
8284 25. When execve() failed while trying to run a command in a pipe transport,
8285 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8286 could be confused with a return value of 69 from the command itself. This
8287 has been changed to 127, the value the shell returns if it is asked to run
8288 a non-existent command. The wording for the related log line suggests a
8289 non-existent command as the problem.
8291 26. If received_header_text expands to an empty string, do not add a Received:
8292 header line to the message. (Well, it adds a token one on the spool, but
8293 marks it "old" so that it doesn't get used or transmitted.)
8295 27. Installed eximstats 1.28 (addition of -nt option).
8297 28. There was no check for failure on the call to getsockname() in the daemon
8298 code. This can fail if there is a shortage of resources on the system, with
8299 ENOMEM, for example. A temporary error is now given on failure.
8301 29. Contrary to the C standard, it seems that in some environments, the
8302 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8303 program. Exim now does this explicitly; it affects the formatting of
8304 timestamps using strftime().
8306 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8307 complaints. I've now initialized all the variables, to avoid this.
8309 32. Header lines added by a system filter were not being "seen" during
8310 transport-time rewrites.
8312 33. The info_callback() function passed to OpenSSL is set up with type void
8313 (*)(SSL *, int, int), as described somewhere. However, when calling the
8314 function (actually a macro) that sets it up, the type void(*)() is
8315 expected. I've put in a cast to prevent warnings from picky compilers.
8317 34. If a DNS black list lookup found a CNAME record, but there were no A
8318 records associated with the domain it pointed at, Exim crashed.
8320 35. If a DNS black list lookup returned more than one A record, Exim ignored
8321 all but the first. It now scans all returned addresses if a particular IP
8322 value is being sought. In this situation, the contents of the
8323 $dnslist_value variable are a list of all the addresses, separated by a
8326 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8327 to accept a host name and all its aliases if the forward lookup for any of
8328 them yielded the IP address of the incoming connection. Now it accepts only
8329 those names whose forward lookup yields the correct IP address. Any other
8330 names are discarded. This closes a loophole whereby a rogue DNS
8331 administrator could create reverse DNS records to break through a
8332 wildcarded host restriction in an ACL.
8334 37. If a user filter or a system filter that ran in a subprocess used any of
8335 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8336 the wrong values were passed to the pipe command ($thisaddress had the
8337 value of $0, $0 had the value of $1, etc). This bug was introduced by
8338 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8340 38. Improved the line breaking for long SMTP error messages from ACLs.
8341 Previously, if there was no break point between 40 and 75 characters, Exim
8342 left the rest of the message alone. Two changes have been made: (a) I've
8343 reduced the minimum length to 35 characters; (b) if it can't find a break
8344 point between 35 and 75 characters, it looks ahead and uses the first one
8345 that it finds. This may give the occasional overlong line, but at least the
8346 remaining text gets split now.
8348 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8349 file descriptors might be low, and that setting 1000 would always raise it.
8350 It turns out that in some environments, the limit is already over 1000 and
8351 that lowering it causes trouble. So now Exim takes care not to decrease it.
8353 40. When delivering a message, the value of $return_path is set to $sender_
8354 address at the start of routing (routers may change the value). By an
8355 oversight, this default was not being set up when an address was tested by
8356 -bt or -bv, which affected the outcome if any router or filter referred to
8359 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8360 log only when "message" or "log_message" are set. However, if one of the
8361 conditions was an address verification, or a call to a nested ACL, the
8362 messages generated by the underlying test were being passed through. This
8363 no longer happens. The underlying message is available in $acl_verify_
8364 message for both "message" and "log_message" expansions, so it can be
8365 passed through if needed.
8367 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8368 new expansion $bh_ to give the encoded byte string without charset
8369 translation. Translation happens only if iconv() is available; HAVE_ICONV
8370 indicates this at build time. HEADERS_CHARSET gives the charset to
8371 translate to; headers_charset can change it in the configuration, and
8372 "headers charset" can change it in an individual filter file.
8374 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8375 that creates RFC 2047 encoded "words" labels them as that charset instead
8376 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8377 expansion operator; (ii) when Exim creates a From: line for a local
8378 message; (iii) when a header line is rewritten to include a "phrase" part.
8380 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8381 buggy, causing it to skip the first lines of messages whose message ID
8382 ended in 'D'. This would not have bitten before Exim release 4.14, because
8383 message IDs were unlikely to end in 'D' before then. The effect was to have
8384 incorrect size information for certain domains.
8386 45. #include "config.h" was missing at the start of the crypt16.c module. This
8387 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8390 46. If there was a timeout during a "random" callout check, Exim treated it as
8391 a failure of the random address, and carried on sending RSET and the real
8392 address. If the delay was just some slowness somewhere, the response to the
8393 original RCPT would be taken as a response to RSET and so on, causing
8394 mayhem of various kinds.
8396 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8397 when I implemented it. It didn't allow for the fact that some option values
8398 may legitimately be negative (e.g. size_addition), and it didn't even do
8399 the right test for positive values.
8401 48. Domain names in DNS records are case-independent. Exim always looks them up
8402 in lower case. Some resolvers return domain names in exactly the case they
8403 appear in the zone file, that is, they may contain uppercase letters. Not
8404 all resolvers do this - some return always lower case. Exim was treating a
8405 change of case by a resolver as a change of domain, similar to a widening
8406 of a domain abbreviation. This triggered its re-routing code and so it was
8407 trying to route what was effectively the same domain again. This normally
8408 caused routing to fail (because the router wouldn't handle the domain
8409 twice). Now Exim checks for this case specially, and just changes the
8410 casing of the domain that it ultimately uses when it transmits the message
8413 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8416 50. If a filter generated a file delivery with a non-absolute name (possible if
8417 no home directory exists for the router), the forbid_file option was not
8420 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8421 the existing equality matching.
8423 52. Exim was using ints instead of ino_t variables in some places where it was
8424 dealing with inode numbers.
8426 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8427 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8428 called TMPDIR, and if it finds it is different, it changes its value.
8430 54. The smtp_printf() function is now made available to local_scan() so
8431 additional output lines can be written before returning. There is also an
8432 smtp_fflush() function to enable the detection of a dropped connection.
8433 The variables smtp_input and smtp_batched_input are exported to
8436 55. Changed the default runtime configuration: the message "Unknown user"
8437 has been removed from the ACL, and instead placed on the localuser router,
8438 using the cannot_route_message feature. This means that any verification
8439 failures that generate their own messages won't get overridden. Similarly,
8440 the "Unrouteable address" message that was in the ACL for unverifiable
8441 relay addresses has also been removed.
8443 56. Added hosts_avoid_esmtp to the smtp transport.
8445 57. The exicyclog script was not checking for the esoteric option
8446 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8447 will work only if exicyclog is run under the appropriate euid.
8449 58. Following a discussion on the list, the rules by which Exim recognises line
8450 endings on incoming messages have been changed. The -dropcr and drop_cr
8451 options are now no-ops, retained only for backwards compatibility. The
8452 following line terminators are recognized: LF CRLF CR. However, special
8453 processing applies to CR:
8455 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8456 nor a local message in the state where . is a terminator.
8458 (ii) If a bare CR is encountered in a header line, an extra space is added
8459 after the line terminator so as not to end the header. The reasoning
8460 behind this is that bare CRs in header lines are most likely either
8461 to be mistakes, or people trying to play silly games.
8463 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8464 was being incorrectly given as 18 bytes larger than it should have been.
8465 This is a VOB (very old bug).
8467 60. This may never have affected anything current, but just in case it has:
8468 When the local host is found other than at the start of a list of hosts,
8469 the local host, those with the same MX, and any that follow, are discarded.
8470 When the list in question was part of a longer list of hosts, the following
8471 hosts (not currently being processed) were also being discarded. This no
8472 longer happens. I'm not sure if this situation could ever has previously
8475 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8478 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8479 header. This is something that is recommended in a new Internet Draft, and
8480 is something that is documented as being done by Sendmail. There are two
8481 possible values. For messages generated by the autoreply transport, Exim
8484 Auto-Submitted: auto-replied
8486 whereas for all other generated messages (e.g. bounces) it adds
8488 Auto-Submitted: auto-generated
8490 63. The "personal" condition in filters now includes a test for the
8491 Auto-Submitted: header. If it contains the string "auto-" the message it
8492 not considered personal.
8494 64. Added rcpt_include_affixes as a generic transport option.
8496 65. Added queue_only_override (default true).
8498 66. Added the syslog_duplication option.
8500 67. If what should have been the first header line of a message consisted of
8501 a space followed by a colon, Exim was mis-interpreting it as a header line.
8502 It isn't of course - it is syntactically invalid and should therefore be
8503 treated as the start of the message body. The misbehaviour could have
8504 caused a number of strange effects, including loss of data in subsequent
8505 header lines, and spool format errors.
8507 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8508 client host had authenticated. This control can now be exercised by an ACL
8509 for more flexibility.
8511 69. By default, callouts do not happen when testing with -bh. There is now a
8512 variant, -bhc, which does actually run the callout code, including
8513 consulting and updating the callout cache.
8515 70. Added support for saslauthd authentication, courtesy of Alexander
8518 71. If statvfs() failed on the spool or log directories while checking their
8519 size for availability, Exim confusingly gave the error "space shortage".
8520 Furthermore, in debugging mode it crashed with a floating point exception.
8521 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8522 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8523 serious problem, Exim now writes to the main and panic logs when this
8524 happens, with details of the failure. It then refuses to accept the
8525 incoming message, giving the message "spool directory problem" or "log
8526 directory problem" with a 421 code for SMTP messages.
8528 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8529 0, 1, and 2 exist, because some OS complain for execs without them (see
8530 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8531 descriptors. However, the code omitted to check that the open succeeded,
8532 causing mysterious errors if for some reason the permissions on /dev/null
8533 got screwed. Now Exim writes a message to the main and panic logs, and
8534 bombs out if it can't open /dev/null.
8536 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8537 interact so that it is all more flexible. It is supposed to remain
8538 backwards compatible. Also added extra_local_interfaces.
8540 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8541 to bomb out with an assertion failure - to the client this appears as a
8542 connection drop. This problem occurs in the part of the code that was taken
8543 from the Samba project. Fortunately, the assertion is in a very simple
8544 function, so I have fixed this by reproducing the function inline in the
8545 one place where it is called, and arranging for authentication to fail
8546 instead of killing the process with assert().
8548 75. The SPA client code was not working when the server requested OEM rather
8549 than Unicode encoding.
8551 76. Added code to make require_files with a specific uid setting more usable in
8552 the case where statting the file as root fails - usually a non-root-mounted
8553 NFS file system. When this happens and the failure is EACCES, Exim now
8554 forks a subprocess and does the per-uid checking as the relevant uid.
8556 77. Added process_log_path.
8558 78. If log_file_path was not explicitly set, a setting of check_log_space or
8559 check_log_inodes was ignored.
8561 79. If a space check for the spool or log partitions fails, the incident is now
8562 logged. Of course, in the latter case the data may get lost...
8564 80. Added the %p formatting code to string_format() so that it can be used to
8565 print addresses in debug_print(). Adjusted all the address printing in the
8566 debugging in store.c to use %p rather than %d.
8568 81. There was a concern that a line of code in smtp_in.c could overflow a
8569 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8570 initially expressed, the concern was not well-founded, because trailing
8571 spaces are removed early. However, if the trailing spaces were followed by
8572 a NULL, they did not get removed, so the overflow was possible. Two fixes
8575 (a) I re-wrote the offending code in a cleaner fashion.
8576 (b) If an incoming SMTP command contains a NULL character, it is rejected
8579 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8580 runs initgroups(), so that if the Exim user is in any additional groups,
8581 they will be used during message reception.
8587 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob