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.
27 JH/04 Fix TLS resumption for TLS-on-connect. This was broken by the advent
28 of loadbalancer-detection for resumption, in 4.96 - which tries to
29 use the EHLO response. SMTPS does not have one at the time it is starting
30 TLS. Change the default for the smtp transport host_name_extract option
31 to be a static string, for TLS-on-connect cases; meaning that resumption
32 will always be attempted (unless deliberately overriden).
39 JH/01 The hosts_connection_nolog main option now also controls "no MAIL in
40 SMTP connection" log lines.
42 JH/02 Option default value updates:
43 - queue_fast_ramp (main) true (was false)
44 - remote_max_parallel (main) 4 (was 2)
46 JH/03 Cache static regex pattern compilations, for use by ACLs.
48 JH/04 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
49 Make the rewrite never match and keep the logging. Trust the
50 admin to be using verify=header-syntax (to actually reject the message).
52 JH/05 Follow symlinks for placing a watch on TLS creds files. This means
53 (under Linux) we watch the dir containing the final file; previously
54 it would be the dir with the first symlink. We still do not monitor
57 JH/06 Check for bad chars in rDNS for sender_host_name. The OpenBSD (at least)
58 dn_expand() is happy to pass them through.
60 JH/07 OpenSSL Fix auto-reload of changed server OCSP proof. Previously, if
61 the file with the proof had an unchanged name, the new proof(s) were
62 loaded on top of the old ones (and nover used; the old ones were stapled).
64 JH/08 Bug 2915: Fix use-after-free for $regex<n> variables. Previously when
65 more than one message arrived in a single connection a reference from
66 the earlier message could be re-used. Often a sigsegv resulted.
67 These variables were introduced in Exim 4.87.
68 Debug help from Graeme Fowler.
70 JH/09 Fix ${filter } for conditions that modify $value. Previously the
71 modified version would be used in construction the result, and a memory
74 JH/10 GnuTLS: fix for (IOT?) clients offering no TLS extensions at all.
75 Find and fix by Jasen Betts.
77 JH/11 OpenSSL: fix for ancient clients needing TLS support for versions earlier
78 than TLSv1,2, Previously, more-recent versions of OpenSSL were permitting
79 the systemwide configuration to override the Exim config.
81 HS/01 Bug 2728: Introduce EDITME option "DMARC_API" to work around incompatible
82 API changes in libopendmarc.
84 JH/12 Bug 2930: Fix daemon startup. When started from any process apart from
85 pid 1, in the normal "background daemon" mode, having to drop process-
86 group leadership also lost track of needing to create listener sockets.
88 JH/13 Bug 2929: Fix using $recipients after ${run...}. A change made for 4.96
89 resulted in the variable appearing empty. Find and fix by Ruben Jenster.
91 JH/14 Bug 2933: Fix regex substring match variables for null matches. Since 4.96
92 a capture group which obtained no text (eg. "(abc)*" matching zero
93 occurrences) could cause a segfault if the corresponding $<n> was
96 JH/15 Fix argument parsing for ${run } expansion. Previously, when an argument
97 included a close-brace character (eg. it itself used an expansion) an
100 JH/16 Move running the smtp connect ACL to before, for TLS-on-connect ports,
101 starting TLS. Previously it was after, meaning that attackers on such
102 ports had to be screened using the host_reject_connection main config
103 option. The new sequence aligns better with the STARTTLS behaviour, and
104 permits defences against crypto-processing load attacks, even though it
105 is strictly an incompatible change.
106 Also, avoid sending any SMTP fail response for either the connect ACL
107 or host_reject_connection, for TLS-on-connect ports.
109 JH/17 Permit the ACL "encrypted" condition to be used in a HELO/EHLO ACL,
110 Previously this was not permitted, but it makes reasonable sense.
111 While there, restore a restriction on using it from a connect ACL; given
112 the change JH/16 it could only return false (and before 4.91 was not
115 JH/18 Fix a fencepost error in logging. Previously (since 4.92) when a log line
116 was exactly sized compared to the log buffer, a crash occurred with the
117 misleading message "bad memory reference; pool not found".
118 Found and traced by Jasen Betts.
120 JH/19 Bug 2911: Fix a recursion in DNS lookups. Previously, if the main option
121 dns_again_means_nonexist included an element causing a DNS lookup which
122 itself returned DNS_AGAIN, unbounded recursion occurred. Possible results
123 included (though probably not limited to) a process crash from stack
124 memory limit, or from excessive open files. Replace this with a paniclog
125 whine (as this is likely a configuration error), and returning
128 JH/20 Bug 2954: (OpenSSL) Fix setting of explicit EC curve/group. Previously
129 this always failed, probably leading to the usual downgrade to in-clear
132 JH/21 Fix TLSA lookups. Previously dns_again_means_nonexist would affect
133 SERVFAIL results, which breaks the downgrade resistance of DANE. Change
134 to not checking that list for these lookups.
136 JH/22 Bug 2434: Add connection-elapsed "D=" element to more connection
139 JH/23 Fix crash in string expansions. Previously, if an empty variable was
140 immediately followed by an expansion operator, a null-indirection read
141 was done, killing the process.
143 JH/24 Bug 2997: When built with EXPERIMENTAL_DSN_INFO, bounce messages can
144 include an SMTP response string which is longer than that supported
145 by the delivering transport. Alleviate by wrapping such lines before
148 JH/25 Bug 2827: Restrict size of References: header in bounce messages to 998
149 chars (RFC limit). Previously a limit of 12 items was made, which with
150 a not-impossible References: in the message being bounced could still
151 be over-large and get stopped in the transport.
153 JH/26 For a ${readsocket } in TLS mode, send a TLS Close Alert before the TCP
154 close. Previously a bare socket close was done.
156 JH/27 Fix ${srs_encode ..}. Previously it would give a bad result for one day
159 JH/28 Bug 2996: Fix a crash in the smtp transport. When finding that the
160 message being considered for delivery was already being handled by
161 another process, and having an SMTP connection already open, the function
162 to close it tried to use an uninitialized variable. This would afftect
163 high-volume sites more, especially when running mailing-list-style loads.
164 Pollution of logs was the major effect, as the other process delivered
165 the message. Found and partly investigated by Graeme Fowler.
167 JH/29 Change format of the internal ID used for message identification. The old
168 version only supported 31 bits for a PID element; the new 64 (on systems
169 which can use Base-62 encoding, which is all currently supported ones
170 but not Darwin (MacOS) or Cygwin, which have case-insensitive filesystems
171 and must use Base-36). The new ID is 23 characters rather than 16, and is
172 visible in various places - notably logs, message headers, and spool file
173 names. Various of the ancillary utilities also have to know the format.
174 As well as the expanded PID portion, the sub-second part of the time
175 recorded in the ID is expanded to support finer precision. Theoretically
176 this permits a receive rate from a single comms channel of better than the
178 The major timestamp part of the ID is not changed; at 6 characters it is
179 usable until about year 3700.
180 Updating from previously releases is fully supported: old-format spool
181 files are still usable, and the utilities support both formats. New
182 message will use the new format. The one hints-DB file type which uses
183 message-IDs (the transport wait- DB) will be discarded if an old-format ID
184 is seen; new ones will be built with only new-format IDs.
185 Optionally, a utility can be used to convert spool files from old to new,
186 but this is only an efficiency measure not a requirement for operation
187 Downgrading from new to old requires running a provided utility, having
188 first stopped all operations. This will convert any spool files from new
189 back to old (losing time-precision and PID information) and remove any
190 wait- hints databases.
192 JH/30 Bug 3006: Fix handling of JSON strings having embedded commas. Previously
193 we treated them as item separators when parsing for a list item, but they
194 need to be protected by the doublequotes. While there, add handling for
197 JH/31 Bug 2998: Fix ${utf8clean:...} to disallow UTF-16 surrogate codepoints.
198 Found and fixed by Jasen Betts. No testcase for this as my usual text
199 editor insists on emitting only valid UTF-8.
201 JH/32 Fix "tls_dhparam = none" under GnuTLS. At least with 3.7.9 this gave
202 a null-indirection SIGSEGV for the receive process.
204 JH/33 Fix free for live variable $value created by a ${run ...} expansion during
205 -bh use. Internal checking would spot this and take a panic.
207 JH/34 Bug 3013: Fix use of $recipients within arguments for ${run...}.
208 In 4.96 this would expand to empty.
210 JH/35 Bug 3014: GnuTLS: fix expiry date for an auto-generated server
211 certificate. Find and fix by Andreas Metzler.
213 JH/36 Add ARC info to DMARC hostory records.
215 JH/37 Bug 3016: Avoid sending DSN when message was accepted under fakereject
216 or fakedefer. Previously the sender could discover that the message
217 had in fact been accepted.
219 JH/38 Taint-track intermediate values from the peer in multi-stage authentation
220 sequences. Previously the input was not noted as being tainted; notably
221 this resulted in behaviour of LOGIN vs. PLAIN being inconsistent under
222 bad coding of authenticators.
224 JH/39 Bug 3023: Fix crash induced by some combinations of zero-length strings
225 and ${tr...}. Found and diagnosed by Heiko Schlichting.
227 JH/40 Bug 2999: Fix a possible OOB write in the external authenticator, which
228 could be triggered by externally-supplied input. Found by Trend Micro.
231 JH/41 Bug 3000: Fix a possible OOB write in the SPA authenticator, which could
232 be triggered by externally-controlled input. Found by Trend Micro.
235 JH/42 Bug 3001: Fix a possible OOB read in the SPA authenticator, which could
236 be triggered by externally-controlled input. Found by Trend Micro.
239 JH/43 Bug 2903: avoid exit on an attempt to rewrite a malformed address.
240 Make the rewrite never match and keep the logging. Trust the
241 admin to be using verify=header-syntax (to actually reject the message).
243 JH/44 Bug 3033: Harden dnsdb lookups against crafted DNS responses.
246 HS/02 Fix string_is_ip_address() CVE-2023-42117 (Bug 3031)
252 JH/01 Move the wait-for-next-tick (needed for unique message IDs) from
253 after reception to before a subsequent reception. This should
254 mean slightly faster delivery, and also confirmation of reception
257 JH/02 Move from using the pcre library to pcre2. The former is no longer
258 being developed or supported (by the original developer).
260 JH/03 Constification work in the filters module required a major version
261 bump for the local-scan API. Specifically, the "headers_charset"
262 global which is visible via the API is now const and may therefore
263 not be modified by local-scan code.
265 JH/04 Fix ClamAV TCP use under FreeBSD. Previously the OS-specific shim for
266 sendfile() didi not account for the way the ClamAV driver code called it.
268 JH/05 Bug 2819: speed up command-line messages being read in. Previously a
269 time check was being done for every character; replace that with one
272 JH/06 Bug 2815: Fix ALPN sent by server under OpenSSL. Previously the string
273 sent was prefixed with a length byte.
275 JH/07 Change the SMTP feature name for pipelining connect to be compliant with
276 RFC 5321. Previously Dovecot (at least) would log errors during
279 JH/08 Remove stripping of the binaries from the FreeBSD build. This was added
280 in 4.61 without a reason logged. Binaries will be bigger, which might
281 matter on diskspace-constrained systems, but debug is easier.
283 JH/09 Fix macro-definition during "-be" expansion testing. The move to
284 write-protected store for macros had not accounted for these runtime
285 additions; fix by removing this protection for "-be" mode.
287 JH/10 Convert all uses of select() to poll(). FreeBSD 12.2 was found to be
288 handing out large-numbered file descriptors, violating the usual Unix
289 assumption (and required by Posix) that the lowest possible number will be
290 allocated by the kernel when a new one is needed. In the daemon, and any
291 child procesees, values higher than 1024 (being bigger than FD_SETSIZE)
292 are not useable for FD_SET() [and hence select()] and overwrite the stack.
293 Assorted crashes happen.
295 JH/11 Fix use of $sender_host_name in daemon process. When used in certain
296 main-section options or in a connect ACL, the value from the first ever
297 connection was never replaced for subsequent connections. Found by
300 JH/12 Bug 2838: Fix for i32lp64 hard-align platforms. Found for SPARC Linux,
301 though only once PCRE2 was introduced: the memory accounting used under
302 debug offset allocations by an int, giving a hard trap in early startup.
303 Change to using a size_t. Debug and fix by John Paul Adrian Glaubitz.
305 JH/13 Bug 2845: Fix handling of tls_require_ciphers for OpenSSL when a value
306 with underbars is given. The write-protection of configuration introduced
307 in 4.95 trapped when normalisation was applied to an option not needing
310 JH/14 Bug 1895: TLS: Deprecate RFC 5114 Diffie-Hellman parameters.
312 JH/15 Fix a resource leak in *BSD. An off-by-one error resulted in the daemon
313 failing to close the certificates directory, every hour or any time it
316 JH/16 Debugging initiated by an ACL control now continues through into routing
317 and transport processes. Previously debugging stopped any time Exim
318 re-execs, or for processing a queued message.
320 JH/17 The "expand" debug selector now gives more detail, specifically on the
321 result of expansion operators and items.
323 JH/18 Bug 2751: Fix include_directory in redirect routers. Previously a
324 bad comparison between the option value and the name of the file to
325 be included was done, and a mismatch was wrongly identified.
326 4.88 to 4.95 are affected.
328 JH/19 Support for Berkeley DB versions 1 and 2 is withdrawn.
330 JH/20 When built with NDBM for hints DB's check for nonexistence of a name
331 supplied as the db file-pair basename. Previously, if a directory
332 path was given, for example via the autoreply "once" option, the DB
333 file.pag and file.dir files would be created in that directory's
336 JH/21 Remove the "allow_insecure_tainted_data" main config option and the
337 "taint" log_selector. These were previously deprecated.
339 JH/22 Fix static address-list lookups to properly return the matched item.
340 Previously only the domain part was returned.
342 JH/23 Bug 2864: FreeBSD: fix transport hang after 4xx/5xx response. Previously
343 the call into OpenSSL to send a TLS Close was being repeated; this
344 resulted in the library waiting for the peer's Close. If that was never
345 sent we waited forever. Fix by tracking send calls.
347 JH/24 The ${run} expansion item now expands its command string elements after
348 splitting. Previously it was before; the new ordering makes handling
349 zero-length arguments simpler. The old ordering can be obtained by
350 appending a new option "preexpand", after a comma, to the "run".
352 JH/25 Taint-check exec arguments for transport-initiated external processes.
353 Previously, tainted values could be used. This affects "pipe", "lmtp" and
354 "queryprogram" transport, transport-filter, and ETRN commands.
355 The ${run} expansion is also affected: in "preexpand" mode no part of
356 the command line may be tainted, in default mode the executable name
359 JH/26 Fix CHUNKING on a continued-transport. Previously the usabliility of
360 the the facility was not passed across execs, and only the first message
361 passed over a connection could use BDAT; any further ones using DATA.
363 JH/27 Support the PIPECONNECT facility in the smtp transport when the helo_data
364 uses $sending_ip_address and an interface is specified.
365 Previously any use of the local address in the EHLO name disabled
366 PIPECONNECT, the common case being to use the rDNS of it.
368 JH/28 OpenSSL: fix transport-required OCSP stapling verification under session
369 resumption. Previously verify failed because no certificate status is
370 passed on the wire for the restarted session. Fix by using the recorded
371 ocsp status of the stored session for the new connection.
373 JH/29 TLS resumption: the key for session lookup in the client now includes
374 more info that a server could potentially use in configuring a TLS
375 session, avoiding oferring mismatching sessions to such a server.
376 Previously only the server IP was used.
378 JH/30 Fix string_copyn() for limit greater than actual string length.
379 Previously the copied amount was the limit, which could result in a
380 overlapping memcpy for newly allocated destination soon after a
381 source string shorter than the limit. Found/investigated by KM.
383 JH/31 Bug 2886: GnuTLS: Do not free the cached creds on transport connection
384 close; it may be needed for a subsequent connection. This caused a
385 SEGV on primary-MX defer. Found/investigated by Gedalya & Andreas.
387 JH/32 Fix CHUNKING for a second message on a connection when the first was
388 rejected. Previously we did not reset the chunking-offered state, and
389 erroneously rejected the BDAT command. Investigation help from
392 JH/33 Fis ${srs_encode ...} to handle an empty sender address, now returning
393 an empty address. Previously the expansion returned an error.
395 HS/01 Bug 2855: Handle a v4mapped sender address given us by a frontending
396 proxy. Previously these were misparsed, leading to paniclog entries.
402 JH/01 Bug 1329: Fix format of Maildir-format filenames to match other mail-
403 related applications. Previously an "H" was used where available info
404 says that "M" should be, so change to match.
406 JH/02 Bug 2587: Fix pam expansion condition. Tainted values are commonly used
407 as arguments, so an implementation trying to copy these into a local
408 buffer was taking a taint-enforcement trap. Fix by using dynamically
409 created buffers. Similar fix for radius expansion condition.
411 JH/03 Bug 2586: Fix listcount expansion operator. Using tainted arguments is
412 reasonable, eg. to count headers. Fix by using dynamically created
413 buffers rather than a local. Do similar fixes for ACL actions "dcc",
414 "log_reject_target", "malware" and "spam"; the arguments are expanded
415 so could be handling tainted values.
417 JH/04 Bug 2590: Fix -bi (newaliases). A previous code rearrangement had
418 broken the (no-op) support for this sendmail command. Restore it
419 to doing nothing, silently, and returning good status.
421 JH/05 Bug 2593: Fix "vacation" in Exim filter. Previously, when a "once"
422 record path was given (or the default used) without a leading directory
423 path, an error occurred on trying to open it. Use the transport's working
426 JH/06 Bug 2594: Change the name used for certificate name checks in the smtp
427 transport. Previously it was the name on the DNS A-record; use instead
428 the head of the CNAME chain leading there (if there is one). This seems
429 to align better with RFC 6125.
431 JH/07 Bug 2597: Fix a resource leak. Using a lookup in obtaining a value for
432 smtp_accept_max_per_host allocated resources which were not released
433 when the limit was exceeded. This eventually crashed the daemon. Fix
434 by adding a release action in that path.
436 JH/08 Bug 2598: Fix verify ACL condition. The options for the condition are
437 expanded; previously using tainted values was rejected. Fix by using
438 dynamically-created buffers.
440 JH/09 Relax restrictions on ACL verify condition needing access to message
441 headers. Previously they were only permitted in data and non-smtp ACLs;
442 permit also mime, dkim, prdr quit and notquit. Applies to header-syntax,
443 not_blind, header_sender and header_names_ascii verification.
445 JH/10 Bug 2603: Fix coding of string copying to only evaluate arguments once.
446 Previously a macro used one argument twice; when called with the
447 argument as an expression having side-effects, incorrect operation
448 resulted. Use an inlineable function.
450 JH/11 Bug 2604: Fix request to cutthrough-deliver when a connection is already
451 held open for a verify callout. Previously this wan not accounted for
452 and a corrupt onward SMTP conversation resulted.
454 JH/12 Bug 2607: Fix the ${srs_encode } expansion to handle quoted local_parts.
455 Previously they were embedded naively in the constructed address; when
456 needed, strip the quoting and quote the entire local_part.
457 Also make the inbound_srs expansion condition handle quoting.
459 JH/13 Fix dsearch "subdir" filter to ignore ".". Previously only ".." was
460 excluded, not matching the documentation.
462 JH/14 Bug 2606: Fix a segfault in sqlite lookups. When no, or a bad, filename
463 was given for the sqlite_dbfile a trap resulted.
465 JH/15 Bug 2620: Fix "spam" ACL condition. Previously, tainted values for the
466 "name" argument resulted in a trap. There is no reason to disallow such;
467 this was a coding error.
469 JH/16 Bug 2615: Fix pause during message reception, on systems that have been
470 suspended/resumed. The Linux CLOCK_MONOTONIC does not account for time
471 spent suspended, ignoring the POSIX definition. Previously we assumed
472 it did and a constant offset from real time could be used as a correction.
473 Change to using the same clock source for the start-of-message and the
474 post-message next-tick-wait. Also change to using CLOCK_BOOTTIME if it
475 exists, just to get a clock slightly more aligned to reality.
477 JH/17 Bug 2295: Fix DKIM signing to always semicolon-terminate. Although the
478 RFC says it is optional some validators care. The missing char was not
479 intended but triggered by a line-wrap alignment. Discovery and fix by
480 Guillaume Outters, hacked on by JH.
482 JH/18 Bug 2617: Fix a taint trap in parse_fix_phrase(). Previously when the
483 name being quoted was tainted a trap would be taken. Fix by using
484 dynamically created buffers. The routine could have been called by a
485 rewrite with the "h" flag, by using the "-F" command-line option, or
486 by using a "name=" option on a control=submission ACL modifier.
488 JH/19 SPF: change the Authentication-Results expansion component to give
489 smtp.helo when the sender domain is empty. Previously it gave
492 JH/20 Bug 2631: ACL dnslist conditions now ignore and log any lookups returns
493 not in 127.0.0.0/8 to help in spotting list domains taken over by a
494 domain-parking registrar.
496 JH/21 Bug 2630: Fix eol-replacement string for the ${readsocket } expansion.
497 Previously when a whitespace character was specified it was not inserted
498 after removing the newline.
500 JH/22 Bug 2265: Force SNI usage for smtp transport DANE'd connections, to be
501 the domain part of the recipient address. This overrides any tls_sni
502 option set, which was previously used.
504 JH/23 Logging: with the +tls_sni log_selector, do not wrap the received SNI
507 JH/24 Bug 2634: Fix a taint trap seen on NetBSD: the testing coded for
508 is_tainted() had an off-by-one error in the overenthusiastic direction.
509 Find and fix by Gavan. Although NetBSD is not a supported platform for
510 4.94 this bug could affect other platforms.
512 PP/01 Fix default prime selection to be consistent.
513 One path used ike23 still, instead of exim.dev.20160529.3; now both
514 execution flows will use the same DH primes (currently
515 exim.dev.20160529.3).
517 JH/25 OpenSSL: Fix back-compatibility behaviour surrounding tls_certificates
518 option in smtp transport, to match the documentation. Previously
519 verification was not being done in some cases where it should have been.
521 JH/26 Bug 2646: fix a memory usage issue in ldap lookups. Previously, when more
522 than one server was defined and depending on the platform memory layout
523 details, an internal consistency trap could be hit while walking the list
526 JH/27 Bug 2648: fix the passing of an authenticator public-name through spool
527 files. The value is used by the authresults expansion item. Previously
528 if this was used in a router or transport, a crash could result.
530 JH/28 Fix spurious logging of select error. Some platforms, notably FreeBSD,
531 have a sufficient incidence of EINTR returns from select that an
532 interaction with other operations done by the main daemon loop exposed
533 a bug in the error-handling. This was benign apart from the log
536 JH/29 Bug 2675: add outgoing-interface I= element to deferred "==" log lines,
537 for consistency with delivered "=>" and failed "**" lines. While we're
538 there, handle PRX and TFO.
540 JH/30 Bug 2677: fix matching of long addresses. Since 4.93 a limit of 256 was
541 applied. This resulted, if any header-line rewrite rules were configured,
542 in a panic-log triggerable by sending a message with a long address in
543 a header. Fix by increasing the arbitrary limit to larger than a single
544 (dewrapped) 5322 header line maximum size.
546 JH/31 The ESMTP option name advertised for the SUPPORT_EARLY_PIPE build option
547 is changed from X_PIPE_CONNECT to PIPE_CONNECT. This is in line with
548 RFC 6648 which deprecates X- options in protocols as a general practice.
549 Changeover between the implementations is handled by the mechanisms
552 JH/32 Bug 2599: fix delay of delivery to a local address where there is also
553 a remote which uses callout/hold. Previously the local was queued.
555 JH/33 Fix a taint trap in the ${listextract } expansion when the source data
558 JH/34 Fix the placement of a multiple-message delivery marker in the delivery
559 log line. The asterisk is now consistently appended to the remote IP
560 (and port, if given), and will also be provided on defer and fail log
561 lines. Previously it could be placed on the local IP if that was being
562 logged, and was only provided on delivery lines.
564 JH/35 Bug 2343: Harden exim_tidydb against corrupt wait- files.
566 JH/36 Bug 2687: Fix interpretation of multiple ^ chars in a plaintext
567 authenticator client_send option. Previously the next char, after a pair
568 was collapsed, was taken verbatim (so ^^^foo became ^^foo; ^^^^foo became
569 ^^\x00foo). Fixed to get ^\x00foo and ^^foo respectively to match the
570 documentation. There is still no way to get a leading ^ immediately
571 after a NUL (ie. for the password of a PLAIN method authenticator.
573 JH/37 Enforce the expected size, for fixed-size records read from hints-DB
574 files. For bad sizes read, delete the record and whine to paniclog.
576 JH/38 When logging an AUTH failure, as server, do not include sensitive
577 information. Previously, the credentials would be included if given
578 as part of the AUTH command line and an ACL denied authentication.
580 JH/39 Bug 2691: fix $local_part_data. When the matching list element
581 referred to a file, bad data was returned. This likely also affected
584 JH/40 The gsasl authenticator now supports caching of the salted password
585 generated by the client-side implementation. This required the addition
586 of a new variable: $auth4.
588 JH/41 Fix daemon SIGHUP on FreeBSD. Previously, a named socket for IPC was
589 left undeleted; the attempt to re-create it then failed - resulting in
590 the usual "SIGHUP tp have daemon reload configuration" to not work.
591 This affected any platform not supporting "abstract" Unix-domain
592 sockets (i.e. not Linux).
594 JH/42 Bug 2693: Harden against a peer which reneges on a 452 "too many
595 recipients" response to RCPT in a later response, with a 250. The
596 previous coding assumed this would not happen, and under PIPELINING
597 would result in both lost and duplicate recipients for a message.
599 JH/43 Bug 2694: Fix weighted distribution of work to multiple spamd servers.
600 Previously the weighting was incorrectly applied. Similar fix for socks
601 proxies. Found and fixed by Heiko Schlichting.
603 JH/44 Bug 2701: Fix list-expansion of dns_ipv4_lookup. Previously, it did
604 not handle sub-lists included using the +namedlist syntax. While
605 investigating, the same found for dns_trust_aa, dns_again_means_nonexist,
606 dnssec_require_domains, dnssec_request_domains, srv_fail_domains,
609 JH/45 Use a (new) separate store pool-pair for DKIM verify working data.
610 Previously the permanent pool was used, so the sore could not be freed.
611 This meant a connection with many messages would use continually-growing
614 JH/46 Use an exponentially-increasing block size when malloc'ing store. Do it
615 per-pool so as not to waste too much space. Previously a constant size
616 was used which resulted in O(n^2) behaviour; now we get O(n log n) making
617 DOS attacks harder. The cost is wasted memory use in the larger blocks.
619 JH/47 Use explicit alloc/free for DNS lookup workspace. This permits using the
620 same space repeatedly, and a smaller process footprint.
622 JH/48 Use a less bogus-looking filename for a temporary used for DH-parameters
623 for GnuTLS. Previously the name started "%s" which, while not a bug,
624 looked as if if might be one.
626 JH/49 Bug 2710: when using SOCKS for additional messages after the first (a
627 "continued connection") make the $proxy_* variables available. Previously
628 the information was not passed across the exec() call for subsequent
629 transport executions. This also mean that the log lines for the
630 messages can show the proxy information.
632 JH/50 Bug 2672: QT elements in log lines, unless disabled, now exclude the
633 receive time. With modern systems the difference is significant.
634 The historical behaviour can be restored by disabling (a new) log_selector
635 "queue_time_exclusive".
637 JH/51 Taint-check ACL line. Previously, only filenames (for out-of-line ACL
638 content) were specifically tested for. Now, also cover expansions
639 resulting in ACL names and inline ACL content.
641 JH/52 Fix ${ip6norm:} operator. Previously, any trailing line text was dropped,
642 making it unusable in complex expressions.
644 JH/53 Bug 2743: fix immediate-delivery via named queue. Previously this would
645 fail with a taint-check on the spoolfile name, and leave the message
648 HS/01 Enforce absolute PID file path name.
650 HS/02 Handle SIGINT as we handle SIGTERM: terminate the Exim process.
652 PP/01 Add a too-many-bad-recipients guard to the default config's RCPT ACL.
654 PP/02 Bug 2643: Correct TLS DH constants.
655 A missing NUL termination in our code-generation tool had led to some
656 incorrect Diffie-Hellman constants in the Exim source.
657 Reported by kylon94, code-gen tool fix by Simon Arlott.
659 PP/03 Impose security length checks on various command-line options.
660 Fixes CVE-2020-SPRSS reported by Qualys.
662 PP/04 Fix Linux security issue CVE-2020-SLCWD and guard against PATH_MAX
663 better. Reported by Qualys.
665 PP/05 Fix security issue CVE-2020-PFPSN and guard against cmdline invoker
666 providing a particularly obnoxious sender full name.
669 PP/06 Fix CVE-2020-28016 (PFPZA): Heap out-of-bounds write in parse_fix_phrase()
671 PP/07 Refuse to allocate too little memory, block negative/zero allocations.
674 PP/08 Change default for recipients_max from unlimited to 50,000.
676 PP/09 Fix security issue with too many recipients on a message (to remove a
677 known security problem if someone does set recipients_max to unlimited,
678 or if local additions add to the recipient list).
679 Fixes CVE-2020-RCPTL reported by Qualys.
681 PP/10 Fix security issue in SMTP verb option parsing
682 Fixes CVE-2020-EXOPT reported by Qualys.
684 PP/11 Fix security issue in BDAT state confusion.
685 Ensure we reset known-good where we know we need to not be reading BDAT
686 data, as a general case fix, and move the places where we switch to BDAT
687 mode until after various protocol state checks.
688 Fixes CVE-2020-BDATA reported by Qualys.
690 HS/03 Die on "/../" in msglog file names
692 QS/01 Creation of (database) files in $spool_dir: only uid=0 or the uid of
693 the Exim runtime user are allowed to create files.
695 QS/02 PID file creation/deletion: only possible if uid=0 or uid is the Exim
698 QS/03 When reading the output from interpreted forward files we do not
699 pass the pipe between the parent and the interpreting process to
700 executed child processes (if any).
702 QS/04 Always die if requested from internal logging, even is logging is
705 JH/54 DMARC: recent versions of the OpenDMARC library appear to have broken
706 the API; compilation noo longer completes with DMARC support included.
707 This affects 1.4.1-1 on Fedora 33 (1.3.2-3 is functional); and has
708 been reported on other platforms.
710 JH/55 TLS: as server, reject connections with ALPN indicating non-smtp use.
712 JH/56 Make the majority of info read from config files readonly, for defence-in-
713 depth against exploits. Suggestion by Qualys.
714 Not supported on Solaris 10.
716 JH/57 Fix control=fakreject for a custom message containing tainted data.
717 Previously this resulted in a log complaint, due to a re-expansion present
718 since fakereject was originally introduced.
720 JH/58 GnuTLS: Fix certextract expansion. If a second modifier after a tag
721 modifier was given, a loop resulted.
723 JH/59 DKIM: Fix small-message verification under TLS with chunking. If a
724 pipelined SMTP command followed the BDAT LAST then it would be
725 incorrectly treated as part of the message body, causing a verification
728 JH/60 Bug 2805: Fix logging of domain-literals in Message_ID: headers. They
729 require looser validation rules than those for 821-level addresses,
730 which only permit IP addresses.
736 JH/01 Avoid costly startup code when not strictly needed. This reduces time
737 for some exim process initialisations. It does mean that the logging
738 of TLS configuration problems is only done for the daemon startup.
740 JH/02 Early-pipelining support code is now included unless disabled in Makefile.
742 JH/03 DKIM verification defaults no long accept sha1 hashes, to conform to
743 RFC 8301. They can still be enabled, using the dkim_verify_hashes main
746 JH/04 Support CHUNKING from an smtp transport using a transport_filter, when
747 DKIM signing is being done. Previously a transport_filter would always
748 disable CHUNKING, falling back to traditional DATA.
750 JH/05 Regard command-line recipients as tainted.
752 JH/06 Bug 340: Remove the daemon pid file on exit, when due to SIGTERM.
754 JH/07 Bug 2489: Fix crash in the "pam" expansion condition. It seems that the
755 PAM library frees one of the arguments given to it, despite the
756 documentation. Therefore a plain malloc must be used.
758 JH/08 Bug 2491: Use tainted buffers for the transport smtp context. Previously
759 on-stack buffers were used, resulting in a taint trap when DSN information
760 copied from a received message was written into the buffer.
762 JH/09 Bug 2493: Harden ARC verify against Outlook, whick has been seen to mix
763 the ordering of its ARC headers. This caused a crash.
765 JH/10 Bug 2492: Use tainted memory for retry record when needed. Previously when
766 a new record was being constructed with information from the peer, a trap
769 JH/11 Bug 2494: Unset the default for dmarc_tld_file. Previously a naiive
770 installation would get error messages from DMARC verify, when it hit the
771 nonexistent file indicated by the default. Distros wanting DMARC enabled
772 should both provide the file and set the option.
773 Also enforce no DMARC verification for command-line sourced messages.
775 JH/12 Fix an uninitialised flag in early-pipelining. Previously connections
776 could, depending on the platform, hang at the STARTTLS response.
778 JH/13 Bug 2498: Reset a counter used for ARC verify before handling another
779 message on a connection. Previously if one message had ARC headers and
780 the following one did not, a crash could result when adding an
781 Authentication-Results: header.
783 JH/14 Bug 2500: Rewind some of the common-coding in string handling between the
784 Exim main code and Exim-related utities. The introduction of taint
785 tracking also did many adjustments to string handling. Since then, eximon
786 frequently terminated with an assert failure.
788 JH/15 When PIPELINING, synch after every hundred or so RCPT commands sent and
789 check for 452 responses. This slightly helps the inefficieny of doing
790 a large alias-expansion into a recipient-limited target. The max_rcpt
791 transport option still applies (and at the current default, will override
792 the new feature). The check is done for either cause of synch, and forces
793 a fast-retry of all 452'd recipients using a new MAIL FROM on the same
794 connection. The new facility is not tunable at this time.
796 JH/16 Fix the variables set by the gsasl authenticator. Previously a pointer to
797 library live data was being used, so the results became garbage. Make
798 copies while it is still usable.
800 JH/17 Logging: when the deliver_time selector ise set, include the DT= field
801 on delivery deferred (==) and failed (**) lines (if a delivery was
802 attemtped). Previously it was only on completion (=>) lines.
804 JH/18 Authentication: the gsasl driver not provides the $authN variables in time
805 for the expansion of the server_scram_iter and server_scram_salt options.
807 WB/01 SPF: DNS lookups for the obsolete SPF RR type done by the libspf2 library
808 are now specifically given a NO_DATA response without hitting the system
809 resolver. The library goes on to do the now-standard TXT lookup.
810 Use of dnsdb lookups is not affected.
812 JH/19 Bug 2507: Modules: on handling a dynamic-module (lookups) open failure,
813 only retrieve the errormessage once. Previously two calls to dlerror()
814 were used, and the second one (for mainlog/paniclog) retrieved null
817 JH/20 Taint checking: disallow use of tainted data for
818 - the appendfile transport file and directory options
819 - the pipe transport command
820 - the autoreply transport file, log and once options
821 - file names used by the redirect router (including filter files)
823 - paths used by single-key lookups
824 Previously this was permitted.
826 JH/21 Bug 2501: Fix init call in the heimdal authenticator. Previously it
827 adjusted the size of a major service buffer; this failed because the
828 buffer was in use at the time. Change to a compile-time increase in the
829 buffer size, when this authenticator is compiled into exim.
831 JH/22 Taint-checking: move to safe-mode taint checking on all platforms. The
832 previous fast-mode was untenable in the face of glibs using mmap to
833 support larger malloc requests.
835 PP/01 Update the openssl_options possible values through OpenSSL 1.1.1c.
836 New values supported, if defined on system where compiled:
837 allow_no_dhe_kex, cryptopro_tlsext_bug, enable_middlebox_compat,
838 no_anti_replay, no_encrypt_then_mac, prioritize_chacha, tlsext_padding
840 JH/23 Performance improvement in the initial phase of a two-pass queue run. By
841 running a limited number of proceses in parallel, a benefit is gained. The
842 amount varies with the platform hardware and load. The use of the option
843 queue_run_in_order means we cannot do this, as ordering becomes
846 JH/24 Bug 2524: fix the cyrus_sasl auth driver gssapi usage. A previous fix
847 had introduced a string-copy (for ensuring NUL-termination) which was not
848 appropriate for that case, which can include embedded NUL bytes in the
849 block of data. Investigation showed the copy to actually be needless, the
850 data being length-specified.
852 JH/25 Fix use of concurrent TLS connections under GnuTLS. When a callout was
853 done during a receiving connection, and both used TLS, global info was
854 used rather than per-connection info for tracking the state of data
855 queued for transmission. This could result in a connection hang.
857 JH/26 Fix use of the SIZE parameter on MAIL commands, on continued connections.
858 Previously, when delivering serveral messages down a single connection
859 only the first would provide a SIZE. This was due to the size information
860 not being properly tracked.
862 JH/27 Bug 2530: When operating in a timezone with sub-minute offset, such as
863 TAI (at 37 seconds currently), pretend to be in UTC for time-related
864 expansion and logging. Previously, spurious values such as a future
865 minute could be seen.
867 JH/28 Bug 2533: Fix expansion of ${tr } item. When called in some situations
868 it could crash from a null-deref. This could also affect the
869 ${addresses: } operator and ${readsock } item.
871 JH/29 Bug 2537: Fix $mime_part_count. When a single connection had a non-mime
872 message following a mime one, the variable was not reset.
874 JH/30 When an pipelined-connect fails at the first response, assume incorrect
875 cached capability (perhaps the peer reneged?) and immediately retry in
878 JH/31 Fix spurious detection of timeout while writing to transport filter.
880 JH/32 Bug 2541: Fix segfault on bad cmdline -f (sender) argument. Previously
881 an attempt to copy the string was made before checking it.
883 JH/33 Fix the dsearch lookup to return an untainted result. Previously the
884 taint of the lookup key was maintained; we now regard the presence in the
885 filesystem as sufficient validation.
887 JH/34 Fix the readsocket expansion to not segfault when an empty "options"
888 argument is supplied.
890 JH/35 The dsearch lookup now requires that the directory is an absolute path.
891 Previously this was not checked, and nonempty relative paths made an
892 access under Exim's current working directory.
894 JH/36 Bug 2554: Fix msg:defer event for the hosts_max_try_hardlimit case.
895 Previously no event was raised.
897 JH/37 Bug 2552: Fix the check on spool space during reception to use the SIZE
898 parameter supplied by the sender MAIL FROM command. Previously it was
899 ignored, and only the check_spool_space option value for the required
902 JH/38 Fix $dkim_key_length. This should, after a DKIM verification, present
903 the size of the signing public-key. Previously it was instead giving
904 the size of the signature hash.
906 JH/39 DKIM verification: the RFC 8301 restriction on sizes of RSA keys is now
907 the default. See the (new) dkim_verify_min_keysizes option.
909 JH/40 Fix a memory-handling bug: when a connection carried multiple messages
910 and an ACL use a lookup for checking either the local_part or domain,
911 stale data could be accessed. Ensure that variable references are
912 dropped between messages.
914 JH/41 Bug 2571: Fix SPA authenticator. Running as a server, an offset supplied
915 by the client was not checked as pointing within response data before
916 being used. A malicious client could thus cause an out-of-bounds read and
917 possibly gain authentication. Fix by adding the check.
919 JH/42 Internationalisation: change the default for downconversion in the smtp
920 transport to be "if needed". Previously it was "as previously set" for
921 the message, which usually meant "if needed" for message-submission but
922 "no" for everything else. However, MTAs have been seen using SMTPUTF8
923 even when the envelope addresses did not need it, resulting in forwarding
924 failures to non-supporting MTAs. A downconvert in such cases will be
925 a no-op on the addresses, merely dropping the use of SMTPUTF8 by the
926 transport. The change does mean that addresses needing conversion will
927 be converted when previously a delivery failure would occur.
929 JH/43 Fix possible long line in DSN. Previously when a very long SMTP error
930 response was received it would be used unchecked in a fail-DSN, violating
931 standards on line-length limits. Truncate if needed.
933 HS/01 Remove parameters of the link to www.open-spf.org. The linked form
934 doesn't work. (Additionally add a new main config option to configure the
941 JH/01 OpenSSL: With debug enabled output keying information sufficient, server
942 side, to decode a TLS 1.3 packet capture.
944 JH/02 OpenSSL: Suppress the sending of (stateful) TLS1.3 session tickets.
945 Previously the default library behaviour applied, sending two, each in
948 JH/03 Debug output for ACL now gives the config file name and line number for
951 JH/04 The default received_header_text now uses the RFC 8314 tls cipher clause.
953 JH/05 DKIM: ensure that dkim_domain elements are lowercased before use.
955 JH/06 Fix buggy handling of autoreply bounce_return_size_limit, and a possible
956 buffer overrun for (non-chunking) other transports.
958 JH/07 GnuTLS: Our use of late (post-handshake) certificate verification, under
959 TLS1.3, means that a server rejecting a client certificate is not visible
960 to the client until the first read of encrypted data (typically the
961 response to EHLO). Add detection for that case and treat it as a failed
962 TLS connection attempt, so that the normal retry-in-clear can work (if
963 suitably configured).
965 JB/01 Bug 2375: fix expansions of 822 addresses having comments in local-part
966 and/or domain. Found and fixed by Jason Betts.
968 JH/08 Add hardening against SRV & TLSA lookups the hit CNAMEs (a nonvalid
969 configuration). If a CNAME target was not a wellformed name pattern, a
972 JH/09 Logging: Fix initial listening-on line for multiple ports for an IP when
973 the OS reports them interleaved with other addresses.
975 JH/10 OpenSSL: Fix aggregation of messages. Previously, when PIPELINING was
976 used both for input and for a verify callout, both encrypted, SMTP
977 responses being sent by the server could be lost. This resulted in
978 dropped connections and sometimes bounces generated by a peer sending
981 JH/11 Harden plaintext authenticator against a badly misconfigured client-send
982 string. Previously it was possible to cause undefined behaviour in a
983 library routine (usually a crash). Found by "zerons".
985 JH/12 Bug 2384: fix "-bP smtp_receive_timeout". Previously it returned no
988 JH/13 Bug 2386: Fix builds with Dane under LibreSSL 2.9.0 onward. Some old
989 API was removed, so update to use the newer ones.
991 JH/14 Bug 1891: Close the log file if receiving a non-smtp message, without
992 any timeout set, is taking a long time. Previously we would hang on to a
993 rotated logfile "forever" if the input was arriving with long gaps
994 (a previous attempt to fix addressed lack, for a long time, of initial
997 HS/01 Bug 2390: Use message_id for tempfile creation to avoid races in a
998 shared (NFS) environment. The length of the tempfile name is now
999 4 + 16 ("hdr.$message_exim_id") which might break on file
1000 systems which restrict the file name length to lower values.
1001 (It was "hdr.$pid".)
1003 HS/02 Bug 2390: Use message_id for tempfile creation to avoid races in a
1004 shared (NFS) environment.
1006 HS/03 Bug 2392: exigrep does case sensitive *option* processing (as it
1007 did for all versions <4.90). Notably -M, -m, --invert, -I may be
1010 JH/15 Use unsigned when creating bitmasks in macros, to avoid build errors
1011 on some platforms for bit 31.
1013 JH/16 GnuTLS: rework ciphersuite strings under recent library versions. Thanks
1014 to changes apparently associated with TLS1.3 handling some of the APIs
1015 previously used were either nonfunctional or inappropriate. Strings
1016 like TLS1.3:ECDHE_SECP256R1__RSA_PSS_RSAE_SHA256__AES_256_GCM__AEAD:256
1017 and TLS1.2:ECDHE_SECP256R1__RSA_SHA256__AES_128_CBC__SHA256:128 replace
1018 the previous TLS1.2:ECDHE_RSA_AES_256_GCM_SHA384:256 .
1019 This affects log line X= elements, the $tls_{in,out}_cipher variables,
1020 and the use of specific cipher names in the encrypted= ACL condition.
1022 JH/17 OpenSSL: the default openssl_options now disables ssl_v3.
1024 JH/18 GnuTLS: fix $tls_out_ocsp under hosts_request_ocsp. Previously the
1025 verification result was not updated unless hosts_require_ocsp applied.
1027 JH/19 Bug 2398: fix listing of a named-queue. Previously, even with the option
1028 queue_list_requires_admin set to false, non-admin users were denied the
1031 JH/20 Bug 2389: fix server advertising of usable certificates, under GnuTLS in
1032 directory-of-certs mode. Previously they were advertised despite the
1035 JH/21 The smtp transport option "hosts_noproxy_tls" is now unset by default.
1036 A single TCP connection by a client will now hold a TLS connection open
1037 for multiple message deliveries, by default. Previously the default was to
1040 JH/22 The smtp transport option "hosts_try_dane" now enables all hosts by
1041 default. If built with the facility, DANE will be used. The facility
1042 SUPPORT_DANE is now enabled in the prototype build Makefile "EDITME".
1044 JH/23 The build default is now for TLS to be included; the SUPPORT_TLS define
1045 is replaced with DISABLE_TLS. Either USE_GNUTLS or (the new) USE_OPENSSL
1046 must be defined and you must still, unless you define DISABLE_TLS, manage
1047 the the include-dir and library-file requirements that go with that
1048 choice. Non-TLS builds are still supported.
1050 JH/24 Fix duplicated logging of peer name/address, on a transport connection-
1053 JH/25 The smtp transport option "hosts_try_fastopen" now enables all hosts by
1054 default. If the platform supports and has the facility enabled, it will
1055 be requested on all coneections.
1057 JH/26 The PIPE_CONNECT facility is promoted from experimental status and is now
1058 controlled by the build-time option SUPPORT_PIPE_CONNECT.
1060 PP/01 Unbreak heimdal_gssapi, broken in 4.92.
1062 JH/27 Bug 2404: Use the main-section configuration option "dsn_from" for
1063 success-DSN messages. Previously the From: header was always the default
1064 one for these; the option was ignored.
1066 JH/28 Fix the timeout on smtp response to apply to the whole response.
1067 Previously it was reset for every read, so a teergrubing peer sending
1068 single bytes within the time limit could extend the connection for a
1069 long time. Credit to Qualsys Security Advisory Team for the discovery.
1071 JH/29 Fix DSN Final-Recipient: field. Previously it was the post-routing
1072 delivery address, which leaked information of the results of local
1073 forwarding. Change to the original envelope recipient address, per
1076 JH/30 Bug 2411: Fix DSN generation when RFC 3461 failure notification is
1077 requested. Previously not bounce was generated and a log entry of
1078 error ignored was made.
1080 JH/31 Avoid re-expansion in ${sort } expansion. (CVE-2019-13917)
1082 JH/32 Introduce a general tainting mechanism for values read from the input
1083 channel, and values derived from them. Refuse to expand any tainted
1084 values, to catch one form of exploit.
1086 JH/33 Bug 2413: Fix dkim_strict option. Previously the expansion result
1087 was unused and the unexpanded text used for the test. Found and
1088 fixed by Ruben Jenster.
1090 JH/34 Fix crash after TLS shutdown. When the TCP/SMTP channel was left open,
1091 an attempt to use a TLS library read routine dereffed a nul pointer,
1094 JH/35 Bug 2409: filter out-of-spec chars from callout response before using
1095 them in our smtp response.
1097 JH/36 Have the general router option retry_use_local_part default to true when
1098 any of the restrictive preconditions are set (to anything). Previously it
1099 was only for check_local user. The change removes one item of manual
1100 configuration which is required for proper retries when a remote router
1101 handles a subset of addresses for a domain.
1103 JH/37 Appendfile: when evaluating quota use (non-quota_size_regex) take the file
1104 link count into consideration.
1106 HS/04 Fix handling of very log lines in -H files. If a -<key> <value> line
1107 caused the extension of big_buffer, the following lines were ignored.
1109 JH/38 Bug 1395: Teach the DNS negative-cache about TTL value from the SOA in
1110 accordance with RFC 2308. Previously there was no expiry, so a longlived
1111 receive process (eg. due to ACL delays) versus a short SOA value could
1114 HS/05 Handle trailing backslash gracefully. (CVE-2019-15846)
1116 JH/39 Promote DMARC support to mainline.
1118 JH/40 Bug 2452: Add a References: header to DSNs.
1120 JH/41 With GnuTLS 3.6.0 (and later) do not attempt to manage Diffie-Hellman
1121 parameters. The relevant library call is documented as "Deprecated: This
1122 function is unnecessary and discouraged on GnuTLS 3.6.0 or later. Since
1123 3.6.0, DH parameters are negotiated following RFC7919."
1125 HS/06 Change the default of dnssec_request_domains to "*"
1127 JH/42 Bug 2545: Fix CHUNKING for all RCPT commands rejected. Previously we
1128 carried on and emitted a BDAT command, even when PIPELINING was not
1131 JH/43 Bug 2465: Fix taint-handling in dsearch lookup. Previously a nontainted
1132 buffer was used for the filename, resulting in a trap when tainted
1133 arguments (eg. $domain) were used.
1135 JH/44 With OpenSSL 1.1.1 (onwards) disable renegotiation for TLS1.2 and below;
1136 recommended to avoid a possible server-load attack. The feature can be
1137 re-enabled via the openssl_options main cofiguration option.
1139 JH/45 local_scan API: documented the current smtp_printf() call. This changed
1140 for version 4.90 - adding a "more data" boolean to the arguments.
1141 Bumped the ABI version number also, this having been missed previously;
1142 release versions 4.90 to 4.92.3 inclusive were effectively broken in
1143 respect of usage of smtp_printf() by either local_scan code or libraries
1144 accessed via the ${dlfunc } expansion item. Both will need coding
1145 adjustment for any calls to smtp_printf() to match the new function
1146 signature; a FALSE value for the new argument is always safe.
1148 JH/46 FreeBSD: fix use of the sendfile() syscall. The shim was not updating
1149 the file-offset (which the Linux syscall does, and exim expects); this
1150 resulted in an indefinite loop.
1152 JH/47 ARC: fix crash in signing, triggered when a configuration error failed
1153 to do ARC verification. The Authentication-Results: header line added
1154 by the configuration then had no ARC item.
1156 JH/48 Bug 2784: fix shutdown=no in the ${readsocket) expansion item. Previously
1157 an incorrect mode was used for reading the result, resulting in it being
1164 JH/01 Remove code calling the customisable local_scan function, unless a new
1165 definition "HAVE_LOCAL_SCAN=yes" is present in the Local/Makefile.
1167 JH/02 Bug 1007: Avoid doing logging from signal-handlers, as that can result in
1168 non-signal-safe functions being used.
1170 JH/03 Bug 2269: When presented with a received message having a stupidly large
1171 number of DKIM-Signature headers, disable DKIM verification to avoid
1172 a resource-consumption attack. The limit is set at twenty.
1174 JH/04 Add variables $arc_domains, $arc_oldest_pass for ARC verify. Fix the
1175 report of oldest_pass in ${authres } in consequence, and separate out
1176 some descriptions of reasons for verification fail.
1178 JH/05 Bug 2273: Cutthrough delivery left a window where the received messsage
1179 files in the spool were present and unlocked. A queue-runner could spot
1180 them, resulting in a duplicate delivery. Fix that by doing the unlock
1181 after the unlink. Investigation by Tim Stewart. Take the opportunity to
1182 add more error-checking on spoolfile handling while that code is being
1185 PP/01 Refuse to open a spool data file (*-D) if it's a symlink.
1186 No known attacks, no CVE, this is defensive hardening.
1188 JH/06 Bug 2275: The MIME ACL unlocked the received message files early, and
1189 a queue-runner could start a delivery while other operations were ongoing.
1190 Cutthrough delivery was a common victim, resulting in duplicate delivery.
1191 Found and investigated by Tim Stewart. Fix by using the open message data
1192 file handle rather than opening another, and not locally closing it (which
1193 releases a lock) for that case, while creating the temporary .eml format
1194 file for the MIME ACL. Also applies to "regex" and "spam" ACL conditions.
1196 JH/07 Bug 177: Make a random-recipient callout success visible in ACL, by setting
1197 $sender_verify_failure/$recipient_verify_failure to "random".
1199 JH/08 When generating a selfsigned cert, use serial number 1 since zero is not
1202 JH/09 Bug 2274: Fix logging of cmdline args when starting in an unlinked cwd.
1203 Previously this would segfault.
1205 JH/10 Fix ARC signing for case when DKIM signing failed. Previously this would
1208 JH/11 Bug 2264: Exim now only follows CNAME chains one step by default. We'd
1209 like zero, since the resolver should be doing this for us, But we need one
1210 as a CNAME but no MX presence gets the CNAME returned; we need to check
1211 that doesn't point to an MX to declare it "no MX returned" rather than
1212 "error, loop". A new main option is added so the older capability of
1213 following some limited number of chain links is maintained.
1215 JH/12 Add client-ip info to non-pass iprev ${authres } lines.
1217 JH/13 For receent Openssl versions (1.1 onward) use modern generic protocol
1218 methods. These should support TLS 1.3; they arrived with TLS 1.3 and the
1219 now-deprecated earlier definitions used only specified the range up to TLS
1220 1.2 (in the older-version library docs).
1222 JH/14 Bug 2284: Fix DKIM signing for body lines starting with a pair of dots.
1224 JH/15 Rework TLS client-side context management. Stop using a global, and
1225 explicitly pass a context around. This enables future use of TLS for
1226 connections to service-daemons (eg. malware scanning) while a client smtp
1227 connection is using TLS; with cutthrough connections this is quite likely.
1229 JH/16 Fix ARC verification to do AS checks in reverse order.
1231 JH/17 Support a "tls" option on the ${readsocket } expansion item.
1233 JH/18 Bug 2287: Fix the protocol name (eg utf8esmtp) for multiple messages
1234 using the SMTPUTF8 option on their MAIL FROM commands, in one connection.
1235 Previously the "utf8" would be re-prepended for every additional message.
1237 JH/19 Reject MAIL FROM commands with SMTPUTF8 when the facility was not advertised.
1238 Previously thery were accepted, resulting in issues when attempting to
1239 forward messages to a non-supporting MTA.
1241 PP/02 Let -n work with printing macros too, not just options.
1243 JH/20 Bug 2296: Fix cutthrough for >1 address redirection. Previously only
1244 one parent address was copied, and bogus data was used at delivery-logging
1245 time. Either a crash (after delivery) or bogus log data could result.
1246 Discovery and analysis by Tim Stewart.
1248 PP/03 Make ${utf8clean:} expansion operator detect incomplete final character.
1249 Previously if the string ended mid-character, we did not insert the
1250 promised '?' replacement.
1252 PP/04 Documentation: current string operators work on bytes, not codepoints.
1254 JH/21 Change as many as possible of the global flags into one-bit bitfields; these
1255 should pack well giving a smaller memory footprint so better caching and
1256 therefore performance. Group the declarations where this can't be done so
1257 that the byte-sized flag variables are not interspersed among pointer
1258 variables, giving a better chance of good packing by the compiler.
1260 JH/22 Bug 1896: Fix the envelope from for DMARC forensic reports to be possibly
1261 non-null, to avoid issues with sites running BATV. Previously reports were
1262 sent with an empty envelope sender so looked like bounces.
1264 JH/23 Bug 2318: Fix the noerror command within filters. It wasn't working.
1265 The ignore_error flag wasn't being returned from the filter subprocess so
1266 was not set for later routers. Investigation and fix by Matthias Kurz.
1268 JH/24 Bug 2310: Raise a msg:fail:internal event for each undelivered recipient,
1269 and a msg:complete for the whole, when a message is manually removed using
1270 -Mrm. Developement by Matthias Kurz, hacked on by JH.
1272 JH/25 Avoid fixed-size buffers for pathnames in DB access. This required using
1273 a "Gnu special" function, asprintf() in the DB utility binary builds; I
1274 hope that is portable enough.
1276 JH/26 Bug 2311: Fix DANE-TA verification under GnuTLS. Previously it was also
1277 requiring a known-CA anchor certificate; make it now rely entirely on the
1278 TLSA as an anchor. Checking the name on the leaf cert against the name
1279 on the A-record for the host is still done for TA (but not for EE mode).
1281 JH/27 Fix logging of proxy address. Previously, a pointless "PRX=[]:0" would be
1282 included in delivery lines for non-proxied connections, when compiled with
1283 SUPPORT_SOCKS and running with proxy logging enabled.
1285 JH/28 Bug 2314: Fire msg:fail:delivery event even when error is being ignored.
1286 Developement by Matthias Kurz, tweaked by JH. While in that bit of code,
1287 move the existing event to fire before the normal logging of message
1288 failure so that custom logging is bracketed by normal logging.
1290 JH/29 Bug 2322: A "fail" command in a non-system filter (file) now fires the
1291 msg:fail:internal event. Developement by Matthias Kurz.
1293 JH/30 Bug 2329: Increase buffer size used for dns lookup from 2k, which was
1294 far too small for todays use of crypto signatures stored there. Go all
1295 the way to the max DNS message size of 64kB, even though this might be
1296 overmuch for IOT constrained device use.
1298 JH/31 Fix a bad use of a copy function, which could be used to pointlessly
1299 copy a string over itself. The library routine is documented as not
1300 supporting overlapping copies, and on MacOS it actually raised a SIGABRT.
1302 JH/32 For main options check_spool_space and check_inode_space, where the
1303 platform supports 64b integers, support more than the previous 2^31 kB
1304 (i.e. more than 2 TB). Accept E, P and T multipliers in addition to
1305 the previous G, M, k.
1307 JH/33 Bug 2338: Fix the cyrus-sasl authenticator to fill in the
1308 $authenticated_fail_id variable on authentication failure. Previously
1311 JH/34 Increase RSA keysize of autogen selfsign cert from 1024 to 2048. RHEL 8.0
1312 OpenSSL didn't want to use such a weak key. Do for GnuTLS also, and for
1313 more-modern GnuTLS move from GNUTLS_SEC_PARAM_LOW to
1314 GNUTLS_SEC_PARAM_MEDIUM.
1316 JH/35 OpenSSL: fail the handshake when SNI processing hits a problem, server
1317 side. Previously we would continue as if no SNI had been received.
1319 JH/36 Harden the handling of string-lists. When a list consisted of a sole
1320 "<" character, which should be a list-separator specification, we walked
1321 off past the nul-terimation.
1323 JH/37 Bug 2341: Send "message delayed" warning MDNs (restricted to external
1324 causes) even when the retry time is not yet met. Previously they were
1325 not, meaning that when (say) an account was over-quota and temp-rejecting,
1326 and multiple senders' messages were queued, only one sender would get
1327 notified on each configured delay_warning cycle.
1329 JH/38 Bug 2351: Log failures to extract envelope addresses from message headers.
1331 JH/39 OpenSSL: clear the error stack after an SSL_accept(). With anon-auth
1332 cipher-suites, an error can be left on the stack even for a succeeding
1333 accept; this results in impossible error messages when a later operation
1336 AM/01 Bug 2359: GnuTLS: repeat lowlevel read and write operations while they
1337 return error codes indicating retry. Under TLS1.3 this becomes required.
1339 JH/40 Fix the feature-cache refresh for EXPERIMENTAL_PIPE_CONNECT. Previously
1340 it only wrote the new authenticators, resulting in a lack of tracking of
1341 peer changes of ESMTP extensions until the next cache flush.
1343 JH/41 Fix the loop reading a message header line to check for integer overflow,
1344 and more-often against header_maxsize. Previously a crafted message could
1345 induce a crash of the recive process; now the message is cleanly rejected.
1347 JH/42 Bug 2366: Fix the behaviour of the dkim_verify_signers option. It had
1348 been totally disabled for all of 4.91. Discovery and fix by "Mad Alex".
1354 GF/01 DEFER rather than ERROR on redis cluster MOVED response.
1355 When redis_servers is set to a list of > 1 element, and the Redis servers
1356 in that list are in cluster configuration, convert the REDIS_REPLY_ERROR
1357 case of MOVED into a DEFER case instead, thus moving the query onto the
1358 next server in the list. For a cluster of N elements, all N servers must
1359 be defined in redis_servers.
1361 GF/02 Catch and remove uninitialized value warning in exiqsumm
1362 Check for existence of @ARGV before looking at $ARGV[0]
1364 JH/01 Replace the store_release() internal interface with store_newblock(),
1365 which internalises the check required to safely use the old one, plus
1366 the allocate and data copy operations duplicated in both (!) of the
1367 extant use locations.
1369 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
1370 modifier. This matches the restriction on the commandline.
1372 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
1373 Previously only the last row was returned.
1375 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
1376 we assumed that tags in the header were well-formed, and parsed the
1377 element content after inspecting only the first char of the tag.
1378 Assumptions at that stage could crash the receive process on malformed
1381 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
1382 While running the DKIM ACL we operate on the Permanent memory pool so that
1383 variables created with "set" persist to the DATA ACL. Also (at any time)
1384 DNS lookups that fail create cache records using the Permanent pool. But
1385 expansions release any allocations made on the current pool - so a dnsdb
1386 lookup expansion done in the DKIM ACL releases the memory used for the
1387 DNS negative-cache, and bad things result. Solution is to switch to the
1388 Main pool for expansions.
1389 While we're in that code, add checks on the DNS cache during store_reset,
1390 active in the testsuite.
1391 Problem spotted, and debugging aided, by Wolfgang Breyha.
1393 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
1394 When none of the hosts presented to a transport match an already-open
1395 connection, close it and proceed with the list. Previously we would
1396 queue the message. Spotted by Lena with Yahoo, probably involving
1399 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
1400 Previously a spurious "250 OK id=" response was appended to the proper
1403 JH/08 The "support for" informational output now, which built with Content
1404 Scanning support, has a line for the malware scanner interfaces compiled
1405 in. Interface can be individually included or not at build time.
1407 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
1408 by the template makefile "src/EDITME". The "STREAM" support for an older
1409 ClamAV interface method is removed.
1411 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
1412 rows affected is given instead).
1414 JH/11 The runtime Berkeley DB library version is now additionally output by
1415 "exim -d -bV". Previously only the compile-time version was shown.
1417 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
1418 SMTP connection. Previously, when one had more recipients than the
1419 first, an abortive onward connection was made. Move to full support for
1420 multiple onward connections in sequence, handling cutthrough connection
1421 for all multi-message initiating connections.
1423 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
1424 routers. Previously, a multi-recipient message would fail to match the
1425 onward-connection opened for the first recipient, and cause its closure.
1427 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
1428 a timeout on read on a GnuTLS initiating connection, resulting in the
1429 initiating connection being dropped. This mattered most when the callout
1430 was marked defer_ok. Fix to keep the two timeout-detection methods
1433 JH/15 Relax results from ACL control request to enable cutthrough, in
1434 unsupported situations, from error to silently (except under debug)
1435 ignoring. This covers use with PRDR, frozen messages, queue-only and
1438 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
1440 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
1441 metadata, resulting in a crash in free().
1443 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
1444 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
1445 Broken also in d185889f4, with init system revamp.
1447 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
1448 Previously we abruptly closed the connection after reading a malware-
1449 found indication; now we go on to read the "scan ok" response line,
1452 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
1453 ACL. Previously, a crash would result.
1455 JH/19 Speed up macro lookups during configuration file read, by skipping non-
1456 macro text after a replacement (previously it was only once per line) and
1457 by skipping builtin macros when searching for an uppercase lead character.
1459 JH/20 DANE support moved from Experimental to mainline. The Makefile control
1460 for the build is renamed.
1462 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
1463 was allocated for every new TLS startup, meaning one per message. Fix
1464 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
1466 JH/22 Bug 2236: When a DKIM verification result is overridden by ACL, DMARC
1467 reported the original. Fix to report (as far as possible) the ACL
1468 result replacing the original.
1470 JH/23 Fix memory leak during multi-message connections using STARTTLS under
1471 OpenSSL. Certificate information is loaded for every new TLS startup,
1472 and the resources needed to be freed.
1474 JH/24 Bug 2242: Fix exim_dbmbuild to permit directoryless filenames.
1476 JH/25 Fix utf8_downconvert propagation through a redirect router. Previously it
1479 JH/26 Bug 2253: For logging delivery lines under PRDR, append the overall
1480 DATA response info to the (existing) per-recipient response info for
1481 the "C=" log element. It can have useful tracking info from the
1482 destination system. Patch from Simon Arlott.
1484 JH/27 Bug 2251: Fix ldap lookups that return a single attribute having zero-
1485 length value. Previously this would segfault.
1487 HS/02 Support Avast multiline protoocol, this allows passing flags to
1488 newer versions of the scanner.
1490 JH/28 Ensure that variables possibly set during message acceptance are marked
1491 dead before release of memory in the daemon loop. This stops complaints
1492 about them when the debug_store option is enabled. Discovered specifically
1493 for sender_rate_period, but applies to a whole set of variables.
1494 Do the same for the queue-runner and queue-list loops, for variables set
1495 from spool message files. Do the same for the SMTP per-message loop, for
1496 certain variables indirectly set in ACL operations.
1498 JH/29 Bug 2250: Fix a longstanding bug in heavily-pipelined SMTP input (such
1499 as a multi-recipient message from a mailinglist manager). The coding had
1500 an arbitrary cutoff number of characters while checking for more input;
1501 enforced by writing a NUL into the buffer. This corrupted long / fast
1502 input. The problem was exposed more widely when more pipelineing of SMTP
1503 responses was introduced, and one Exim system was feeding another.
1504 The symptom is log complaints of SMTP syntax error (NUL chars) on the
1505 receiving system, and refused recipients seen by the sending system
1506 (propating to people being dropped from mailing lists).
1507 Discovered and pinpointed by David Carter.
1509 JH/30 The (EXPERIMENTAL_DMARC) variable $dmarc_ar_header is withdrawn, being
1510 replaced by the ${authresults } expansion.
1512 JH/31 Bug 2257: Fix pipe transport to not use a socket-only syscall.
1514 HS/03 Set a handler for SIGTERM and call exit(3) if running as PID 1. This
1515 allows proper process termination in container environments.
1517 JH/32 Bug 2258: Fix spool_wireformat in combination with LMTP transport.
1518 Previously the "final dot" had a newline after it; ensure it is CR,LF.
1520 JH/33 SPF: remove support for the "spf" ACL condition outcome values "err_temp"
1521 and "err_perm", deprecated since 4.83 when the RFC-defined words
1522 "temperror" and "permerror" were introduced.
1524 JH/34 Re-introduce enforcement of no cutthrough delivery on transports having
1525 transport-filters or DKIM-signing. The restriction was lost in the
1526 consolidation of verify-callout and delivery SMTP handling.
1527 Extend the restriction to also cover ARC-signing.
1529 JH/35 Cutthrough: for a final-dot response timeout (and nonunderstood responses)
1530 in defer=pass mode supply a 450 to the initiator. Previously the message
1533 PP/02 DANE: add dane_require_tls_ciphers SMTP Transport option; if unset,
1534 tls_require_ciphers is used as before.
1536 HS/03 Malware Avast: Better match the Avast multiline protocol. Add
1537 "pass_unscanned". Only tmpfails from the scanner are written to
1538 the paniclog, as they may require admin intervention (permission
1539 denied, license issues). Other scanner errors (like decompression
1540 bombs) do not cause a paniclog entry.
1542 JH/36 Fix reinitialisation of DKIM logging variable between messages.
1543 Previously it was possible to log spurious information in receive log
1546 JH/37 Bug 2255: Revert the disable of the OpenSSL session caching. This
1547 triggered odd behaviour from Outlook Express clients.
1549 PP/03 Add util/renew-opendmarc-tlds.sh script for safe renewal of public
1552 JH/38 DKIM: accept Ed25519 pubkeys in SubjectPublicKeyInfo-wrapped form,
1553 since the IETF WG has not yet settled on that versus the original
1554 "bare" representation.
1556 JH/39 Fix syslog logging for syslog_timestamp=no and log_selector +millisec.
1557 Previously the millisecond value corrupted the output.
1558 Fix also for syslog_pid=no and log_selector +pid, for which the pid
1559 corrupted the output.
1565 JH/01 Rework error string handling in TLS interface so that the caller in
1566 more cases is responsible for logging. This permits library-sourced
1567 string to be attached to addresses during delivery, and collapses
1568 pairs of long lines into single ones.
1570 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
1571 during configuration. Wildcards are allowed and expanded.
1573 JH/02 Rework error string handling in DKIM to pass more info back to callers.
1574 This permits better logging.
1576 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
1577 do not close it down and have the child transport start it up again on
1578 the passed-on TCP connection. Instead, proxy the child (and any
1579 subsequent ones) for TLS via a unix-domain socket channel. Logging is
1580 affected: the continued delivery log lines do not have any DNSSEC, TLS
1581 Certificate or OCSP information. TLS cipher information is still logged.
1583 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
1584 identical IP addresses on different listening ports. Will also affect
1587 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
1588 add noisy ifdef guards to special-case this sillyness.
1589 Patch from Bernd Kuhls.
1591 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
1592 than 255 are no longer allowed.
1594 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
1595 Disable the session-cache too, which might reduce our load. Since we
1596 currrectly use a new context for every connection, both as server and
1597 client, there is no benefit for these.
1598 GnuTLS appears to not support tickets server-side by default (we don't
1599 call gnutls_session_ticket_enable_server()) but client side is enabled
1600 by default on recent versions (3.1.3 +) unless the PFS priority string
1603 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
1604 <https://reproducible-builds.org/specs/source-date-epoch/>.
1606 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
1607 the check for any unsuccessful recipients did not notice the limit, and
1608 erroneously found still-pending ones.
1610 JH/08 Pipeline CHUNKING command and data together, on kernels that support
1611 MSG_MORE. Only in-clear (not on TLS connections).
1613 JH/09 Avoid using a temporary file during transport using dkim. Unless a
1614 transport-filter is involved we can buffer the headers in memory for
1615 creating the signature, and read the spool data file once for the
1616 signature and again for transmission.
1618 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
1619 4.77 as the kernel support then wasn't solid, having issues in 64bit
1620 mode. Now, it's been long enough. Add support for FreeBSD also.
1622 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
1623 case where the routing stage had gathered several addresses to send to
1624 a host before calling the transport for the first, we previously failed
1625 to close down TLS in the old transport process before passing the TCP
1626 connection to the new process. The new one sent a STARTTLS command
1627 which naturally failed, giving a failed delivery and bloating the retry
1628 database. Investigation and fix prototype from Wolfgang Breyha.
1630 JH/12 Fix check on SMTP command input synchronisation. Previously there were
1631 false-negatives in the check that the sender had not preempted a response
1632 or prompt from Exim (running as a server), due to that code's lack of
1633 awareness of the SMTP input buffering.
1635 PP/04 Add commandline_checks_require_admin option.
1636 Exim drops privileges sanely, various checks such as -be aren't a
1637 security problem, as long as you trust local users with access to their
1638 own account. When invoked by services which pass untrusted data to
1639 Exim, this might be an issue. Set this option in main configuration
1640 AND make fixes to the calling application, such as using `--` to stop
1643 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
1644 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
1645 responses to those, into a single TLS record each way (this usually means
1646 a single packet). As a side issue, smtp_enforce_sync now works on TLS
1649 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
1650 affects you only if you're dancing at the edge of the param size limits.
1651 If you are, and this message makes sense to you, then: raise the
1652 configured limit or use OpenSSL 1.1. Nothing we can do for older
1655 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
1656 cmdline element to get the documented default one. Previously it was
1659 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
1660 are unsuable for non-message-specific reasons - eg. connection timeout,
1661 banner-time rejection.
1663 JH/16 Fix logging of delivery remote port, when specified by router, under
1666 PP/06 Repair manualroute's ability to take options in any order, even if one
1667 is the name of a transport.
1670 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
1672 JH/17 Change the list-building routines interface to use the expanding-string
1673 triplet model, for better allocation and copying behaviour.
1675 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
1676 Previously it was constructed the first time a possibly-matching string
1677 was met in the configuration file input during startup; now it is done
1680 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
1681 compatible one, to avoid the (poorly documented) possibility of a config
1682 file in the working directory redirecting the DB files, possibly correpting
1683 some existing file. CVE-2017-10140 assigned for BDB.
1685 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
1686 cache-hot. Previously, although the result was properly cached, the
1687 initial verify call returned a defer.
1689 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
1690 the main verify for receipient in uncached-mode.
1692 JH/22 Retire historical build files to an "unsupported" subdir. These are
1693 defined as "ones for which we have no current evidence of testing".
1695 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
1696 if present. Previously it was ignored.
1698 JH/24 Start using specified-initialisers in C structure init coding. This is
1699 a C99 feature (it's 2017, so now considered safe).
1701 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
1702 if was a fixed-sized field and bitmask ops via macros; it is now more
1705 PP/07 GitHub PR 56: Apply MariaDB build fix.
1706 Patch provided by Jaroslav Škarvada.
1708 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
1709 during Coverity cleanups [4.87 JH/47]
1710 Diagnosis and fix provided by Michael Fischer v. Mollard.
1712 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
1713 the right size to place the terminating semicolon on its own folded
1714 line, the header hash was calculated to an incorrect value thanks to
1715 the (relaxed) space the fold became.
1717 HS/02 Fix Bug 2130: large writes from the transport subprocess were chunked
1718 and confused the parent.
1720 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
1721 which could crash as a result. This could lead to undeliverable messages.
1723 JH/28 Logging: "next input sent too soon" now shows where input was truncated
1726 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
1727 matters on fast-turnover and PID-randomising systems, which were getting
1728 out-of-order delivery.
1730 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
1731 a possibly-overlapping copy. The symptom was that "Remote host closed
1732 connection in response to HELO" was logged instead of the actual 4xx
1735 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
1736 Previously only that bufferd was discarded, resulting in SYMTP command
1739 JH/32 DKIM: when a message has multiple signatures matching an identity given
1740 in dkim_verify_signers, run the dkim acl once for each. Previously only
1741 one run was done. Bug 2189.
1743 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
1744 "panic the current process" to "deliberately defer". The panic log is
1745 still written with the problem list name; the mail and reject logs now
1746 get a temp-reject line for the message that was being handled, saying
1747 something like "domains check lookup or other defer". The SMTP 451
1748 message is still "Temporary local problem".
1750 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
1751 A crafted sequence of BDAT commands could result in in-use memory beeing
1752 freed. CVE-2017-16943.
1754 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
1755 from SMTP input. Previously it was always done; now only done for DATA
1756 and not BDAT commands. CVE-2017-16944.
1758 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
1759 to the message (such as an overlong header line). Previously this was
1760 not done and we did not exit BDAT mode. Followon from the previous item
1761 though a different problem.
1767 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
1768 than -2003 did; needs libidn2 in addition to libidn.
1770 JH/02 The path option on a pipe transport is now expanded before use.
1772 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
1773 Patch provided by "Björn", documentation fix added too.
1775 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
1776 missing a wire-to-host endian conversion.
1778 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
1779 close after a BDAT command line could be taken as a following command,
1780 giving a synch failure. Fix by only checking for synch immediately
1781 before acknowledging the chunk.
1783 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
1784 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
1785 macro. Patches provided by Josh Soref.
1787 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
1788 Previously we did not; the RFC seems ambiguous and VRFY is not listed
1789 by IANA as a service extension. However, John Klensin suggests that we
1792 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
1793 the dkim code may be unix-mode line endings rather than smtp wire-format
1794 CRLF, so prepend a CR to any bare LF.
1796 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
1797 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
1799 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
1800 the first were themselves being wrongly included in the feed into dkim
1801 processing; with most chunk sizes in use this resulted in an incorrect
1802 body hash calculated value.
1804 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
1805 DKIM signature block, for verification. Although advised against by
1806 standards it is specifically not ruled illegal.
1808 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
1810 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
1811 missing a body hash (the bh= tag).
1813 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
1814 It seems that HAProxy sends the Proxy Protocol information in clear and
1815 only then does a TLS startup, so do the same.
1817 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
1818 TCP connections (such as for Spamd) unless the daemon successfully set
1819 Fast Open mode on its listening sockets. This fixes breakage seen on
1820 too-old kernels or those not configured for Fast Open, at the cost of
1821 requiring both directions being enabled for TFO, and TFO never being used
1822 by non-daemon-related Exim processes.
1824 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
1825 endings, at least on the first header line. Try to canonify any that get
1826 past that check, despite the cost.
1828 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
1829 now limited to an arbitrary five deep, while parsing addresses with the
1830 strip_excess_angle_brackets option enabled.
1832 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
1833 instead leave the unprompted TLS handshake in socket buffer for the
1834 TLS library to consume.
1836 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
1838 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
1840 JH/16 Drop variables when they go out of scope. Memory management drops a whole
1841 region in one operation, for speed, and this leaves assigned pointers
1842 dangling. Add checks run only under the testsuite which checks all
1843 variables at a store-reset and panics on a dangling pointer; add code
1844 explicitly nulling out all the variables discovered. Fixes one known
1845 bug: a transport crash, where a dangling pointer for $sending_ip_address
1846 originally assigned in a verify callout, is re-used.
1848 PP/06 Drop '.' from @INC in various Perl scripts.
1850 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
1852 PP/08 Reduce a number of compilation warnings under clang; building with
1853 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
1854 should be warning-free.
1856 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
1858 HS/01 Fix portability problems introduced by PP/08 for platforms where
1859 realloc(NULL) is not equivalent to malloc() [SunOS et al].
1861 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
1862 chunk. This allows us to accept broken chunked messages. We need a more
1863 general solution here.
1865 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
1866 already-broken messages in the queue.
1868 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
1870 JH/19 Fix reference counting bug in routing-generated-address tracking.
1876 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
1877 supports it and a size is available (ie. the sending peer gave us one).
1879 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
1880 years of being deprecated). The replacements are the ACLs
1881 acl_smtp_mime and acl_not_smtp_mime.
1883 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
1884 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
1885 or even an in-clear connection were permitted. Now, if the host lookup
1886 was dnssec and dane was requested then the host is only used if the
1887 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
1888 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
1889 if one fails this test.
1890 This means that a poorly-configured remote DNS will make it incommunicado;
1891 but it protects against a DNS-interception attack on it.
1893 JH/04 Bug 1810: make continued-use of an open smtp transport connection
1894 non-noisy when a race steals the message being considered.
1896 JH/05 If main configuration option tls_certificate is unset, generate a
1897 self-signed certificate for inbound TLS connections.
1899 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
1900 in rewrites and routers.
1902 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
1903 and logged a warning sing 4.83; now they are a configuration file error.
1905 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
1906 (lacking @domain). Apply the same qualification processing as RCPT.
1908 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
1910 JH/10 Support ${sha256:} applied to a string (as well as the previous
1913 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
1914 a cutthrough deliver is pending, as we always want to make a connection.
1915 This also avoids re-routing the message when later placing the cutthrough
1916 connection after a verify cache hit.
1917 Do not update it with the verify result either.
1919 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
1920 when routing results in more than one destination address.
1922 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
1923 signing (which inhibits the cutthrough capability). Previously only
1924 the presence of an option was tested; now an expansion evaluating as
1925 empty is permissible (obviously it should depend only on data available
1926 when the cutthrough connection is made).
1928 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
1929 the relevant preceding SMTP command did not note the pipelining mode.
1931 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
1932 Previously they were not counted.
1934 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
1935 as one having no matching records. Previously we deferred the message
1936 that needed the lookup.
1938 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
1939 distinguished as "(=".
1941 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
1942 for missing MX records. Previously it only worked for missing A records.
1944 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
1946 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
1947 after the data-go-ahead and data-ack. Patch from Jason Betts.
1949 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
1950 even for a "none" policy. Patch from Tony Meyer.
1952 JH/22 Fix continued use of a connection for further deliveries. If a port was
1953 specified by a router, it must also match for the delivery to be
1956 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
1957 When one of the recipients of a message was unsuitable for the connection
1958 (has no matching addresses), we lost track of needing to mark it
1959 deferred. As a result mail would be lost.
1961 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
1963 JH/25 Decoding ACL controls is now done using a binary search; the source code
1964 takes up less space and should be simpler to maintain. Merge the ACL
1965 condition decode tables also, with similar effect.
1967 JH/26 Fix problem with one_time used on a redirect router which returned the
1968 parent address unchanged. A retry would see the parent address marked as
1969 delivered, so not attempt the (identical) child. As a result mail would
1972 JH/27 Fix a possible security hole, wherein a process operating with the Exim
1973 UID can gain a root shell. Credit to http://www.halfdog.net/ for
1974 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
1977 JH/28 Enable {spool,log} filesystem space and inode checks as default.
1978 Main config options check_{log,spool}_{inodes,space} are now
1979 100 inodes, 10MB unless set otherwise in the configuration.
1981 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
1982 Previously it only applied to the main-section connection policy
1985 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
1987 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
1988 by me. Added RFC7919 DH primes as an alternative.
1990 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
1991 are not in the system include path.
1993 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
1994 GnuTLS, when a session startup failed (eg because the client disconnected)
1995 Exim did stdio operations after fclose. This was exposed by a recent
1996 change which nulled out the file handle after the fclose.
1998 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
1999 signed directly by the cert-signing cert, rather than an intermediate
2000 OCSP-signing cert. This is the model used by LetsEncrypt.
2002 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
2004 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
2005 an incoming connection.
2007 HS/02 Bug 1802: Do not half-close the connection after sending a request
2010 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
2011 fallback to "prime256v1".
2013 JH/34 SECURITY: Use proper copy of DATA command in error message.
2014 Could leak key material. Remotely exploitable. CVE-2016-9963.
2020 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
2021 and 3.4.4 - once the server is enabled to respond to an OCSP request
2022 it does even when not requested, resulting in a stapling non-aware
2023 client dropping the TLS connection.
2025 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
2026 support variable-length bit vectors. No functional change.
2028 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
2029 The I= interface field on outgoing lines is now after the H= remote
2030 host field, same as incoming lines. There is a separate
2031 outgoing_interface log selector which allows you to disable the
2034 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
2035 If not running log_selector +smtp_connection the mainlog would be held
2036 open indefinitely after a "too many connections" event, including to a
2037 deleted file after a log rotate. Leave the per net connection logging
2038 leaving it open for efficiency as that will be quickly detected by the
2039 check on the next write.
2041 HS/01 Bug 1671: Fix post transport crash.
2042 Processing the wait-<transport> messages could crash the delivery
2043 process if the message IDs didn't exist for some reason. When
2044 using 'split_spool_directory=yes' the construction of the spool
2045 file name failed already, exposing the same netto behaviour.
2047 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
2048 mime_regex ACL conditions.
2050 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
2051 to DSN fail messages (bounces): remote IP, remote greeting, remote response
2052 to HELO, local diagnostic string.
2054 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
2055 log line to debug. Even when configured with a tls authenticator many
2056 client connections are expected to not authenticate in this way, so
2057 an authenticate fail is not an error.
2059 HS/02 Add the Exim version string to the process info. This way exiwhat
2060 gives some more detail about the running daemon.
2062 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
2063 matter for fast-change records such as DNSBLs.
2065 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
2066 retry record, even if constant. There may be multiple transports with
2067 different interface settings and the retry behaviour needs to be kept
2070 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
2072 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
2074 JH/10 Bug 840: fix log_defer_output option of pipe transport
2076 JH/11 Bug 830: use same host for all RCPTS of a message, even under
2077 hosts_randomize. This matters a lot when combined with mua_wrapper.
2079 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
2080 ${quote_pgsql:<string>} operator.
2082 JH/13 Bug 1708: avoid misaligned access in cached lookup.
2084 JH/14 Change header file name for freeradius-client. Relevant if compiling
2085 with Radius support; from the Gentoo tree and checked under Fedora.
2087 JH/15 Bug 1712: Introduce $prdr_requested flag variable
2089 JH/16 Bug 1714: Permit an empty string as expansion result for transport
2090 option transport_filter, meaning no filtering.
2092 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
2094 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
2095 defaults to "*" (all hosts). The variable is now available when not built
2096 with TLS, default unset, mainly to enable keeping the testsuite sane.
2097 If a server certificate is not supplied (via tls_certificate) an error is
2098 logged, and clients will find TLS connections fail on startup. Presumably
2099 they will retry in-clear.
2100 Packagers of Exim are strongly encouraged to create a server certificate
2101 at installation time.
2103 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
2104 with the $config_file variable.
2106 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
2107 in transport context, after the attempt, and per-recipient. The latter type
2108 is per host attempted. The event data is the error message, and the errno
2109 information encodes the lookup type (A vs. MX) used for the (first) host,
2110 and the trailing two digits of the smtp 4xx response.
2112 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
2113 to write to mainlog (or rejectlog, paniclog) in the window between file
2114 creation and permissions/ownership being changed. Particularly affects
2115 installations where exicyclog is run as root, rather than exim user;
2116 result is that the running daemon panics and dies.
2118 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
2120 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
2121 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
2122 "pri" and "weight". Note that the previous implicit priority given by the
2123 list order is no longer honoured.
2125 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
2126 for DKIM processing.
2128 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
2129 by defining SUPPORT_SOCKS.
2131 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
2132 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
2133 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
2134 variables are renamed to proxy_{local,external}_{address,port}.
2136 JH/27 Move Internationalisation support from Experimental to mainline, enabled
2137 for a build by defining SUPPORT_I18N
2139 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
2140 of the query string, and make ${quote_redis:} do that quoting.
2142 JH/29 Move Events support from Experimental to mainline, enabled by default
2143 and removable for a build by defining DISABLE_EVENT.
2145 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
2147 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
2148 cached by the daemon.
2150 JH/32 Move Redis support from Experimental to mainline, enabled for a build
2151 by defining LOOKUP_REDIS. The libhiredis library is required.
2153 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
2154 keys are given for lookup.
2156 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
2157 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
2158 only supported when built with TLS support. The PolarSSL SHA routines
2159 are still used when the TLS library is too old for convenient support.
2161 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
2162 openssl_options), for security. OpenSSL forces this from version 1.1.0
2163 server-side so match that on older versions.
2165 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
2166 allocation for $value could be released as the expansion processing
2167 concluded, but leaving the global pointer active for it.
2169 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
2170 and to use the domains and local_parts ACL conditions.
2172 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
2173 incorrectly not doubled on cutthrough transmission, hence seen as a
2174 body-termination at the receiving system - resulting in truncated mails.
2175 Commonly the sender saw a TCP-level error, and retransmitted the message
2176 via the normal store-and-forward channel. This could result in duplicates
2177 received - but deduplicating mailstores were liable to retain only the
2178 initial truncated version.
2180 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
2182 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
2184 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
2185 we're in there, support oversigning also; bug 1309.
2187 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
2189 HS/04 Add support for keep_environment and add_environment options.
2191 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
2192 either intentional arithmetic overflow during PRNG, or testing config-
2195 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
2196 delivery resulted in actual delivery. Cancel cutthrough before DATA
2199 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
2200 rejects a recipient: pass the reject to the originator.
2202 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
2203 Many were false-positives and ignorable, but it's worth fixing the
2206 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
2207 for the new environment-manipulation done at startup. Move the routines
2208 from being local to tls.c to being global via the os.c file.
2210 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
2211 an extract embedded as result-arg for a map, the first arg for extract
2212 is unavailable so we cannot tell if this is a numbered or keyed
2213 extraction. Accept either.
2219 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
2222 JH/02 The smtp transport option "multi_domain" is now expanded.
2224 JH/03 The smtp transport now requests PRDR by default, if the server offers
2227 JH/04 Certificate name checking on server certificates, when exim is a client,
2228 is now done by default. The transport option tls_verify_cert_hostnames
2229 can be used to disable this per-host. The build option
2230 EXPERIMENTAL_CERTNAMES is withdrawn.
2232 JH/05 The value of the tls_verify_certificates smtp transport and main options
2233 default to the word "system" to access the system default CA bundle.
2234 For GnuTLS, only version 3.0.20 or later.
2236 JH/06 Verification of the server certificate for a TLS connection is now tried
2237 (but not required) by default. The verification status is now logged by
2238 default, for both outbound TLS and client-certificate supplying inbound
2241 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
2244 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
2245 Status Notification (bounce) messages are now MIME format per RFC 3464.
2246 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
2247 under the control of the dsn_advertise_hosts option, and routers may
2248 have a dsn_lasthop option.
2250 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
2251 default, modifiable by a malware= option. The list separator for
2252 the options can now be changed in the usual way. Bug 68.
2254 JH/10 The smtp_receive_timeout main option is now expanded before use.
2256 JH/11 The incoming_interface log option now also enables logging of the
2257 local interface on delivery outgoing connections.
2259 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
2260 if the interface and destination host and port all match.
2262 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
2265 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
2266 Patch from Andrew Lewis.
2268 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
2269 now supports optional time-restrictions, weighting, and priority
2270 modifiers per server. Patch originally by <rommer@active.by>.
2272 JH/16 The spamd_address main option now supports a mixed list of local
2273 and remote servers. Remote servers can be IPv6 addresses, and
2274 specify a port-range.
2276 JH/17 Bug 68: The spamd_address main option now supports an optional
2277 timeout value per server.
2279 JH/18 Bug 1581: Router and transport options headers_add/remove can
2280 now have the list separator specified.
2282 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
2285 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
2288 JH/21 Support for the A6 type of dns record is withdrawn.
2290 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
2291 rather than the verbs used.
2293 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
2294 from 255 to 1024 chars.
2296 JH/24 Verification callouts now attempt to use TLS by default.
2298 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
2299 are generic router options now. The defaults didn't change.
2301 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
2302 Original patch from Alexander Shikoff, worked over by JH.
2304 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
2305 files if ZCAT_COMMAND is not executable.
2307 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
2309 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
2311 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
2312 Normally benign, it bites when the pair was led to by a CNAME;
2313 modern usage is to not canonicalize the domain to a CNAME target
2314 (and we were inconsistent anyway for A-only vs AAAA+A).
2316 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
2318 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
2319 when evaluating $sender_host_dnssec.
2321 JH/31 Check the HELO verification lookup for DNSSEC, adding new
2322 $sender_helo_dnssec variable.
2324 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
2326 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
2328 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
2330 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
2331 documented as working, but never had. Support all but $spam_report.
2333 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
2334 added for tls authenticator.
2336 HS/03 Add perl_taintmode main config option
2342 TL/01 When running the test suite, the README says that variables such as
2343 no_msglog_check are global and can be placed anywhere in a specific
2344 test's script, however it was observed that placement needed to be near
2345 the beginning for it to behave that way. Changed the runtest perl
2346 script to read through the entire script once to detect and set these
2347 variables, reset to the beginning of the script, and then run through
2348 the script parsing/test process like normal.
2350 TL/02 The BSD's have an arc4random API. One of the functions to induce
2351 adding randomness was arc4random_stir(), but it has been removed in
2352 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
2353 function when detected.
2355 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
2356 cause callback expansion.
2358 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
2359 syntax errors in an expansion can be treated as a string instead of
2360 logging or causing an error, due to the internal use of bool_lax
2361 instead of bool when processing it.
2363 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
2364 server certificates when making smtp deliveries.
2366 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
2368 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
2370 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
2372 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
2373 Merged patch from Sebastian Wiedenroth.
2375 JH/05 Fix results-pipe from transport process. Several recipients, combined
2376 with certificate use, exposed issues where response data items split
2377 over buffer boundaries were not parsed properly. This eventually
2378 resulted in duplicates being sent. This issue only became common enough
2379 to notice due to the introduction of connection certificate information,
2380 the item size being so much larger. Found and fixed by Wolfgang Breyha.
2382 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
2383 size buffer was used, resulting in syntax errors when an expansion
2386 JH/07 Add support for directories of certificates when compiled with a GnuTLS
2387 version 3.3.6 or later.
2389 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
2390 is EXPERIMENTAL_EVENT, the main-configuration and transport options
2391 both become "event_action", the variables become $event_name, $event_data
2392 and $event_defer_errno. There is a new variable $verify_mode, usable in
2393 routers, transports and related events. The tls:cert event is now also
2394 raised for inbound connections, if the main configuration event_action
2397 TL/06 In test suite, disable OCSP for old versions of openssl which contained
2398 early OCSP support, but no stapling (appears to be less than 1.0.0).
2400 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
2401 server certificate names available under the smtp transport option
2402 "tls_verify_cert_hostname" now do not permit multi-component wildcard
2405 JH/10 Time-related extraction expansions from certificates now use the main
2406 option "timezone" setting for output formatting, and are consistent
2407 between OpenSSL and GnuTLS compilations. Bug 1541.
2409 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
2410 encoded parameter in the incoming message. Bug 1558.
2412 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
2413 include certificate info, eximon was claiming there were spoolfile
2416 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
2418 JH/14 Log delivery-related information more consistently, using the sequence
2419 "H=<name> [<ip>]" wherever possible.
2421 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
2422 are problematic for Debian distribution, omit them from the release
2425 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
2427 JH/16 Fix string representation of time values on 64bit time_t architectures.
2430 JH/17 Fix a null-indirection in certextract expansions when a nondefault
2431 output list separator was used.
2436 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
2437 checkers that were complaining about end of non-void function with no
2440 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
2441 This was a regression introduced in 4.83 by another bugfix.
2443 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
2445 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
2446 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
2452 TF/01 Correctly close the server side of TLS when forking for delivery.
2454 When a message was received over SMTP with TLS, Exim failed to clear up
2455 the incoming connection properly after forking off the child process to
2456 deliver the message. In some situations the subsequent outgoing
2457 delivery connection happened to have the same fd number as the incoming
2458 connection previously had. Exim would try to use TLS and fail, logging
2459 a "Bad file descriptor" error.
2461 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
2462 utilities have not been installed.
2464 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
2465 temporary space as the ACL may create new global variables.
2467 TL/01 LDAP support uses per connection or global context settings, depending
2468 upon the detected version of the libraries at build time.
2470 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
2471 to extract and use the src ip:port in logging and expansions as if it
2472 were a direct connection from the outside internet. PPv2 support was
2473 updated based on HAProxy spec change in May 2014.
2475 JH/02 Add ${listextract {number}{list}{success}{fail}}.
2477 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
2478 Properly escape header and check for NULL return.
2480 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
2483 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
2485 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
2486 characters in header names, implemented as a verify condition.
2487 Contributed by Michael Fischer v. Mollard.
2489 TL/05 Rename SPF condition results err_perm and err_temp to standardized
2490 results permerror and temperror. Previous values are deprecated but
2491 still accepted. In a future release, err_perm and err_temp will be
2492 completely removed, which will be a backward incompatibility if the
2493 ACL tests for either of these two old results. Patch contributed by
2494 user bes-internal on the mailing list.
2496 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
2498 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
2499 selectors, in both main and reject logs.
2501 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
2504 JH/07 Add malware type "sock" for talking to simple daemon.
2506 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
2508 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
2509 routers/transports under cutthrough routing.
2511 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
2512 numbers. Touch up "bool" conditional to keep the same definition.
2514 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
2516 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
2518 JH/12 Expand items in router/transport headers_add or headers_remove lists
2519 individually rather than the list as a whole. Bug 1452.
2521 Required for reasonable handling of multiple headers_ options when
2522 they may be empty; requires that headers_remove items with embedded
2523 colons must have them doubled (or the list-separator changed).
2525 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
2526 view the policy declared in the DMARC record. Currently, $dmarc_status
2527 is a combined value of both the record presence and the result of the
2530 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
2532 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
2533 dnslookup router and the smtp transport (applying to the forward
2536 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
2537 of ldap servers used for a specific lookup. Patch provided by Heiko
2540 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
2541 New variable $lookup_dnssec_authenticated for observability.
2543 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
2544 Patch submitted by Lars Timman.
2546 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
2548 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
2549 Requires trusted mode and valid format message id, aborts otherwise.
2550 Patch contributed by Heiko Schlichting.
2552 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
2553 certextract with support for various fields. Bug 1358.
2555 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
2556 is requested by default, modifiable by smtp transport option
2559 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
2560 operate on certificate variables to give certificate fingerprints
2561 Also new ${sha256:cert_variable}.
2563 JH/23 The PRDR feature is moved from being Experimental into the mainline.
2565 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
2566 Christian Aistleitner.
2568 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
2570 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
2571 file. Patch from Wolfgang Breyha.
2573 JH/25 Expand the coverage of the delivery $host and $host_address to
2574 client authenticators run in verify callout. Bug 1476.
2576 JH/26 Port service names are now accepted for tls_on_connect_ports, to
2577 align with daemon_smtp_ports. Bug 72.
2579 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
2580 support and error reporting did not work properly.
2582 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
2583 and is readable. Patch from Andrew Colin Kissa.
2585 TL/14 Enhance documentation of ${run expansion and how it parses the
2586 commandline after expansion, particularly in the case when an
2587 unquoted variable expansion results in an empty value.
2589 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
2591 PP/02 Fix internal collision of T_APL on systems which support RFC3123
2592 by renaming away from it. Addresses GH issue 15, reported by
2595 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
2597 TL/15 SECURITY: prevent double expansion in math comparison functions
2598 (can expand unsanitized data). Not remotely exploitable.
2605 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
2607 PP/02 Make -n do something, by making it not do something.
2608 When combined with -bP, the name of an option is not output.
2610 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
2613 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
2614 $sender_host_name and config options to manage this, and basic check
2617 PP/05 DSCP support for outbound connections and control modifier for inbound.
2619 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
2620 (Only plugin which currently uses this is kerberos4, which nobody should
2621 be using, but we should make it available and other future plugins might
2622 conceivably use it, even though it would break NAT; stuff *should* be
2623 using channel bindings instead).
2625 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
2626 name; added for Sendmail compatibility; requires admin caller.
2627 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
2628 just ignore it); requires trusted caller.
2629 Also parse but ignore: -Ac -Am -X<logfile>
2632 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
2634 TL/02 Add +smtp_confirmation as a default logging option.
2636 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
2637 Patch by Magnus Holmgren from 2007-02-20.
2639 TL/04 Bugzilla 1281 - Spec typo.
2640 Bugzilla 1283 - Spec typo.
2641 Bugzilla 1290 - Spec grammar fixes.
2643 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
2645 TL/06 Add Experimental DMARC support using libopendmarc libraries.
2647 TL/07 Fix an out of order global option causing a segfault. Reported to dev
2648 mailing list by by Dmitry Isaikin.
2650 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
2652 JH/02 Support "G" suffix to numbers in ${if comparisons.
2654 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
2656 NM/01 Bugzilla 1197 - Spec typo
2657 Bugzilla 1196 - Spec examples corrections
2659 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
2661 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
2662 gnutls_enable_pkcs11, but renamed to more accurately indicate its
2665 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
2666 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
2668 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
2669 "acl {{name}{arg}...}", and optional args on acl condition
2672 JH/05 Permit multiple router/transport headers_add/remove lines.
2674 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
2676 JH/07 Avoid using a waiting database for a single-message-only transport.
2677 Performance patch from Paul Fisher. Bugzilla 1262.
2679 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
2682 JH/09 Add $headers_added variable, with content from use of ACL modifier
2683 add_header (but not yet added to the message). Bugzilla 199.
2685 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
2686 Pulled from Bugzilla 817 by Wolfgang Breyha.
2688 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
2690 (nb: this is the same fix as in Exim 4.80.1)
2692 JH/11 Add A= logging on delivery lines, and a client_set_id option on
2695 JH/12 Add optional authenticated_sender logging to A= and a log_selector
2698 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
2700 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
2701 advertise SMTP AUTH mechanism to us, instead of a generic
2702 protocol violation error. Also, make Exim more robust to bad
2703 data from the Dovecot auth socket.
2705 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
2707 When a queue runner is handling a message, Exim first routes the
2708 recipient addresses, during which it prunes them based on the retry
2709 hints database. After that it attempts to deliver the message to
2710 any remaining recipients. It then updates the hints database using
2713 So if a recipient address works intermittently, it can get repeatedly
2714 deferred at routing time. The retry hints record remains fresh so the
2715 address never reaches the final cutoff time.
2717 This is a fairly common occurrence when a user is bumping up against
2718 their storage quota. Exim had some logic in its local delivery code
2719 to deal with this. However it did not apply to per-recipient defers
2720 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
2722 This change adds a proper retry rule check during routing so that the
2723 final cutoff time is checked against the message's age. We only do
2724 this check if there is an address retry record and there is not a
2725 domain retry record; this implies that previous attempts to handle
2726 the address had the retry_use_local_parts option turned on. We use
2727 this as an approximation for the destination being like a local
2728 delivery, as in LMTP.
2730 I suspect this new check makes the old local delivery cutoff check
2731 redundant, but I have not verified this so I left the code in place.
2733 TF/02 Correct gecos expansion when From: is a prefix of the username.
2735 Test 0254 submits a message to Exim with the header
2739 When I ran the test suite under the user fanf2, Exim expanded
2740 the header to contain my full name, whereas it should have added
2741 a Resent-Sender: header. It erroneously treats any prefix of the
2742 username as equal to the username.
2744 This change corrects that bug.
2746 GF/01 DCC debug and logging tidyup
2747 Error conditions log to paniclog rather than rejectlog.
2748 Debug lines prefixed by "DCC: " to remove any ambiguity.
2750 TF/03 Avoid unnecessary rebuilds of lookup-related code.
2752 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
2753 Bug spotted by Jeremy Harris; was flawed since initial commit.
2754 Would have resulted in OCSP responses post-SNI triggering an Exim
2755 NULL dereference and crash.
2757 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
2759 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
2760 Bug detection, analysis and fix by Samuel Thibault.
2761 Bugzilla 1331, Debian bug #698092.
2763 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
2765 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
2766 Server implementation by Todd Lyons, client by JH.
2767 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
2768 config variable "prdr_enable" controls whether the server
2769 advertises the facility. If the client requests PRDR a new
2770 acl_data_smtp_prdr ACL is called once for each recipient, after
2771 the body content is received and before the acl_smtp_data ACL.
2772 The client is controlled by both of: a hosts_try_prdr option
2773 on the smtp transport, and the server advertisement.
2774 Default client logging of deliveries and rejections involving
2775 PRDR are flagged with the string "PRDR".
2777 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
2778 fclose(). Diagnosis by Todd Lyons.
2780 PP/17 Update configure.default to handle IPv6 localhost better.
2781 Patch by Alain Williams (plus minor tweaks).
2784 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
2785 This is now consistent with GnuTLS, and is now documented: the
2786 previous undocumented portable approach to treating the option as
2787 unset was to force an expansion failure. That still works, and
2788 an empty string is now equivalent.
2790 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
2791 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
2792 not performing validation itself.
2794 PP/20 Added force_command boolean option to pipe transport.
2795 Patch from Nick Koston, of cPanel Inc.
2797 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
2800 TF/04 Added udpsend ACL modifier and hexquote expansion operator
2802 PP/21 Fix eximon continuous updating with timestamped log-files.
2803 Broken in a format-string cleanup in 4.80, missed when I repaired the
2804 other false fix of the same issue.
2805 Report and fix from Heiko Schlichting.
2808 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
2809 Report from Prashanth Katuri.
2811 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
2812 It's SecureTransport, so affects any MacOS clients which use the
2813 system-integrated TLS libraries, including email clients.
2815 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
2816 using a MIME ACL for non-SMTP local injection.
2817 Report and assistance in diagnosis by Warren Baker.
2819 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
2821 JH/16 Fix comparisons for 64b. Bugzilla 1385.
2823 TL/09 Add expansion variable $authenticated_fail_id to keep track of
2824 last id that failed so it may be referenced in subsequent ACL's.
2826 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
2829 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
2830 ldap library initialization, allowing self-signed CA's to be
2831 used. Also properly sets require_cert option later in code by
2832 using NULL (global ldap config) instead of ldap handle (per
2833 session). Bug diagnosis and testing by alxgomz.
2835 TL/12 Enhanced documentation in the ratelimit.pl script provided in
2836 the src/util/ subdirectory.
2838 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
2839 renamed to Transport Post Delivery Action by Jeremy Harris, as
2842 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
2843 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
2844 redis_servers = needs to be configured which will be used by the redis
2845 lookup. Patch from Warren Baker, of The Packet Hub.
2847 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
2849 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
2850 hostname or reverse DNS when processing a host list. Used suggestions
2851 from multiple comments on this bug.
2853 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
2855 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
2856 Missed a few lines, added it to make the runtest require no keyboard
2859 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
2860 contains upper case chars. Make router use caseful_local_part.
2862 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
2863 support when GnuTLS has been built with p11-kit.
2869 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
2871 This, or similar/improved, will also be change PP/11 of 4.82.
2877 PP/01 Handle short writes when writing local log-files.
2878 In practice, only affects FreeBSD (8 onwards).
2879 Bugzilla 1053, with thanks to Dmitry Isaikin.
2881 NM/01 Bugzilla 949 - Documentation tweak
2883 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
2886 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
2888 PP/02 Implemented gsasl authenticator.
2890 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
2892 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
2893 `pkg-config foo` for cflags/libs.
2895 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
2896 with rest of GSASL and with heimdal_gssapi.
2898 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
2899 `pkg-config foo` for cflags/libs for the TLS implementation.
2901 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
2902 properties get this fed in as external SSF. A number of robustness
2903 and debugging improvements to the cyrus_sasl authenticator.
2905 PP/08 cyrus_sasl server now expands the server_realm option.
2907 PP/09 Bugzilla 1214 - Log authentication information in reject log.
2908 Patch by Jeremy Harris.
2910 PP/10 Added dbmjz lookup type.
2912 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
2914 PP/12 MAIL args handles TAB as well as SP, for better interop with
2915 non-compliant senders.
2916 Analysis and variant patch by Todd Lyons.
2918 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
2919 Bug report from Lars Müller <lars@samba.org> (via SUSE),
2920 Patch from Dirk Mueller <dmueller@suse.com>
2922 PP/13 tls_peerdn now print-escaped for spool files.
2923 Observed some $tls_peerdn in wild which contained \n, which resulted
2924 in spool file corruption.
2926 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
2927 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
2928 or write after TLS renegotiation, which otherwise led to messages
2931 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
2932 as a tracking header (ie: a signed header comes before the signature).
2933 Patch from Wolfgang Breyha.
2935 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
2936 comma-sep list; embedded commas doubled.
2938 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
2940 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
2942 Report and patch from Dmitry Banschikov.
2944 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
2945 Removed SSL_clear() after SSL_new() which led to protocol negotiation
2946 failures. We appear to now support TLS1.1+ with Exim.
2948 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
2949 lets Exim select keys and certificates based upon TLS SNI from client.
2950 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
2951 before an outbound SMTP session. New log_selector, +tls_sni.
2953 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
2954 NULL dereference. Report and patch from Alun Jones.
2956 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
2957 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
2958 Not seeing resolver debug output on NetBSD, but suspect this is a
2959 resolver implementation change.
2961 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
2962 Left warnings. Added "eximon gdb" invocation mode.
2964 PP/21 Defaulting "accept_8bitmime" to true, not false.
2966 PP/22 Added -bw for inetd wait mode support.
2968 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
2969 locate the relevant includes and libraries. Made this the default.
2971 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
2972 Bugzilla 1246, report and most of solution from Tomasz Kusy.
2974 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
2975 This may cause build issues on older platforms.
2977 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
2978 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
2979 gnutls_require_mac & gnutls_require_protocols (no longer supported).
2980 Added SNI support via GnuTLS too.
2981 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
2983 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
2985 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
2986 Applied second patch from Janne, implementing suggestion to default
2987 multiple-strings-in-record handling to match SPF spec.
2989 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
2991 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
2992 read-only, out of scope).
2993 Patch from Wolfgang Breyha, report from Stuart Northfield.
2995 PP/29 Fix three issues highlighted by clang analyser static analysis.
2996 Only crash-plausible issue would require the Cambridge-specific
2997 iplookup router and a misconfiguration.
2998 Report from Marcin Mirosław.
3000 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
3002 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
3003 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
3004 As part of this, removing so much warning spew let me fix some minor
3005 real issues in debug logging.
3007 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
3008 assignment on my part. Fixed.
3010 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
3011 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
3012 Janne Snabb (who went above and beyond: thank you).
3014 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
3015 string otherwise requires a connection and a bunch more work and it's
3016 relatively easy to get wrong. Should also expose TLS library linkage
3019 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
3020 64-bit ${eval} (JH/03).
3022 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
3023 GNU libc to support some of the 64-bit stuff, should not lead to
3024 conflicts. Defined before os.h is pulled in, so if a given platform
3025 needs to override this, it can.
3027 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
3028 protection layer was required, which is not implemented.
3029 Bugzilla 1254, patch from Wolfgang Breyha.
3031 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
3032 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
3033 tls_dhparam take prime identifiers. Also unbreak combination of
3034 OpenSSL+DH_params+TLSSNI.
3036 PP/39 Disable SSLv2 by default in OpenSSL support.
3042 PP/01 Solaris build fix for Oracle's LDAP libraries.
3043 Bugzilla 1109, patch from Stephen Usher.
3045 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
3047 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
3050 TF/02 Fix a couple more cases where we did not log the error message
3051 when unlink() failed. See also change 4.74-TF/03.
3053 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
3054 lock up or crash if it happened to be inside a call to libc when it
3055 got a SIGUSR1 from exiwhat.
3057 The SIGUSR1 handler appends the current process status to the process
3058 log which is later printed by exiwhat. It used to use the general
3059 purpose logging code to do this, but several functions it calls are
3060 not safe for signals.
3062 The new output code in the SIGUSR1 handler is specific to the process
3063 log, and simple enough that it's easy to inspect for signal safety.
3064 Removing some special cases also simplifies the general logging code.
3065 Removing the spurious timestamps from the process log simplifies
3068 TF/04 Improved ratelimit ACL condition.
3070 The /noupdate option has been deprecated in favour of /readonly which
3071 has clearer semantics. The /leaky, /strict, and /readonly update modes
3072 are mutually exclusive. The update mode is no longer included in the
3073 database key; it just determines when the database is updated. (This
3074 means that when you upgrade Exim will forget old rate measurements.)
3076 Exim now checks that the per_* options are used with an update mode that
3077 makes sense for the current ACL. For example, when Exim is processing a
3078 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
3079 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
3080 must specify per_mail/readonly. If you omit the update mode it defaults to
3081 /leaky where that makes sense (as before) or /readonly where required.
3083 The /noupdate option is now undocumented but still supported for
3084 backwards compatibility. It is equivalent to /readonly except that in
3085 ACLs where /readonly is required you may specify /leaky/noupdate or
3086 /strict/noupdate which are treated the same as /readonly.
3088 A useful new feature is the /count= option. This is a generalization
3089 of the per_byte option, so that you can measure the throughput of other
3090 aggregate values. For example, the per_byte option is now equivalent
3091 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
3093 The per_rcpt option has been generalized using the /count= mechanism
3094 (though it's more complicated than the per_byte equivalence). When it is
3095 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
3096 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
3097 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
3098 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
3099 Note that using per_rcpt with a non-readonly update mode in more than
3100 one ACL will cause the recipients to be double-counted. (The per_mail
3101 and per_byte options don't have this problem.)
3103 The handling of very low rates has changed slightly. If the computed rate
3104 is less than the event's count (usually one) then this event is the first
3105 after a long gap. In this case the rate is set to the same as this event's
3106 count, so that the first message of a spam run is counted properly.
3108 The major new feature is a mechanism for counting the rate of unique
3109 events. The new per_addr option counts the number of different
3110 recipients that someone has sent messages to in the last time period. It
3111 behaves like per_rcpt if all the recipient addresses are different, but
3112 duplicate recipient addresses do not increase the measured rate. Like
3113 the /count= option this is a general mechanism, so the per_addr option
3114 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
3115 example, measure the rate that a client uses different sender addresses
3116 with the options per_mail/unique=$sender_address. There are further
3117 details in the main documentation.
3119 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
3121 TF/06 Removed a few PCRE remnants.
3123 TF/07 Automatically extract Exim's version number from tags in the git
3124 repository when doing development or release builds.
3126 PP/02 Raise smtp_cmd_buffer_size to 16kB.
3127 Bugzilla 879. Patch from Paul Fisher.
3129 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
3130 Heavily based on revision 40f9a89a from Simon Arlott's tree.
3133 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
3135 PP/05 Variable $av_failed, true if the AV scanner deferred.
3136 Bugzilla 1078. Patch from John Horne.
3138 PP/06 Stop make process more reliably on build failure.
3139 Bugzilla 1087. Patch from Heiko Schlittermann.
3141 PP/07 Make maildir_use_size_file an _expandable_ boolean.
3142 Bugzilla 1089. Patch from Heiko Schlittermann.
3144 PP/08 Handle ${run} returning more data than OS pipe buffer size.
3145 Bugzilla 1131. Patch from Holger Weiß.
3147 PP/09 Handle IPv6 addresses with SPF.
3148 Bugzilla 860. Patch from Wolfgang Breyha.
3150 PP/10 GnuTLS: support TLS 1.2 & 1.1.
3152 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
3155 PP/11 match_* no longer expand right-hand-side by default.
3156 New compile-time build option, EXPAND_LISTMATCH_RHS.
3157 New expansion conditions, "inlist", "inlisti".
3159 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
3161 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
3163 PP/14 fix log_write() format string regression from TF/03.
3164 Bugzilla 1152. Patch from Dmitry Isaikin.
3170 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
3172 PP/02 Harmonised TLS library version reporting; only show if debugging.
3173 Layout now matches that introduced for other libraries in 4.74 PP/03.
3175 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
3177 PP/04 New "dns_use_edns0" global option.
3179 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
3182 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
3183 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
3185 TK/01 Updated PolarSSL code to 0.14.2.
3186 Bugzilla 1097. Patch from Andreas Metzler.
3188 PP/07 Catch divide-by-zero in ${eval:...}.
3189 Fixes bugzilla 1102.
3191 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
3194 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
3195 format-string attack -- SECURITY: remote arbitrary code execution.
3197 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
3198 time unintentionally subject to list matching rules, letting the header
3199 cause arbitrary Exim lookups (of items which can occur in lists, *not*
3200 arbitrary string expansion). This allowed for information disclosure.
3202 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
3203 INT_MIN/-1 -- value coerced to INT_MAX.
3209 NM/01 Workaround for PCRE version dependency in version reporting
3212 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
3213 This fixes portability to compilers other than gcc, notably
3214 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
3216 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
3217 makefiles for portability to HP-UX and POSIX correctness.
3219 PP/01 Permit LOOKUP_foo enabling on the make command-line.
3220 Also via indented variable definition in the Makefile.
3221 (Debugging by Oliver Heesakkers).
3223 PP/02 Restore caching of spamd results with expanded spamd_address.
3224 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
3226 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
3227 Improves build reliability. Fix from: Frank Elsner
3229 NM/02 Fix wide character breakage in the rfc2047 coding
3230 Fixes bug 1064. Patch from Andrey N. Oktyabrski
3232 NM/03 Allow underscore in dnslist lookups
3233 Fixes bug 1026. Patch from Graeme Fowler
3235 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
3236 Code patches from Adam Ciarcinski of NetBSD.
3238 NM/04 Fixed exiqgrep to cope with mailq missing size issue
3241 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
3242 is logged, to avoid truncation. Patch from John Horne.
3244 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
3245 Patch from Jakob Hirsch.
3247 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
3248 SQL string expansion failure details.
3249 Patch from Andrey Oktyabrski.
3251 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
3252 Patch from Simon Arlott.
3254 PP/09 New lookups functionality failed to compile on old gcc which rejects
3255 extern declarations in function scope.
3256 Patch from Oliver Fleischmann
3258 PP/10 Use sig_atomic_t for flags set from signal handlers.
3259 Check getgroups() return and improve debugging.
3260 Fixed developed for diagnosis in bug 927 (which turned out to be
3263 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
3264 Patch from Mark Zealey.
3266 PP/12 Bugzilla 1056: Improved spamd server selection.
3267 Patch from Mark Zealey.
3269 PP/13 Bugzilla 1086: Deal with maildir quota file races.
3270 Based on patch from Heiko Schlittermann.
3272 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
3273 Patch from Uwe Doering, sign-off by Michael Haardt.
3275 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
3276 variable declaration deep within a block. Bug and patch from
3279 PP/15 lookups-Makefile IRIX compatibility coercion.
3281 PP/16 Make DISABLE_DKIM build knob functional.
3283 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
3284 Patch by Simon Arlott
3286 TF/03 Fix valgrind.h portability to C89 compilers that do not support
3287 variable argument macros. Our copy now differs from upstream.
3293 TF/01 Failure to get a lock on a hints database can have serious
3294 consequences so log it to the panic log.
3296 TF/02 Log LMTP confirmation messages in the same way as SMTP,
3297 controlled using the smtp_confirmation log selector.
3299 TF/03 Include the error message when we fail to unlink a spool file.
3301 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
3302 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
3303 for maintaining out-of-tree patches for some time.
3305 PP/01 Bugzilla 139: Documentation and portability issues.
3306 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
3307 Handle per-OS dynamic-module compilation flags.
3309 PP/02 Let /dev/null have normal permissions.
3310 The 4.73 fixes were a little too stringent and complained about the
3311 permissions on /dev/null. Exempt it from some checks.
3312 Reported by Andreas M. Kirchwitz.
3314 PP/03 Report version information for many libraries, including
3315 Exim version information for dynamically loaded libraries. Created
3316 version.h, now support a version extension string for distributors
3317 who patch heavily. Dynamic module ABI change.
3319 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
3320 privilege escalation vulnerability whereby the Exim run-time user
3321 can cause root to append content of the attacker's choosing to
3324 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
3327 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
3328 If dropping privileges for untrusted macros, we disabled normal logging
3329 on the basis that it would fail; for the Exim run-time user, this is not
3330 the case, and it resulted in successful deliveries going unlogged.
3331 Fixed. Reported by Andreas Metzler.
3337 PP/01 Date: & Message-Id: revert to normally being appended to a message,
3338 only prepend for the Resent-* case. Fixes regression introduced in
3339 Exim 4.70 by NM/22 for Bugzilla 607.
3341 PP/02 Include check_rfc2047_length in configure.default because we're seeing
3342 increasing numbers of administrators be bitten by this.
3344 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
3346 PP/03 Bugzilla 994: added openssl_options main configuration option.
3348 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
3350 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
3352 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
3354 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
3355 without a peer certificate, leading to a segfault because of an
3356 assumption that peers always have certificates. Be a little more
3357 paranoid. Problem reported by Martin Tscholak.
3359 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
3360 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
3361 NB: ClamAV planning to remove STREAM in "middle of 2010".
3362 CL also introduces -bmalware, various -d+acl logging additions and
3363 more caution in buffer sizes.
3365 PP/09 Implemented reverse_ip expansion operator.
3367 PP/10 Bugzilla 937: provide a "debug" ACL control.
3369 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
3371 PP/12 Bugzilla 973: Implement --version.
3373 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
3375 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
3377 PP/15 Bugzilla 816: support multiple condition rules on Routers.
3379 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
3380 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
3381 ignore trailing whitespace.
3383 JJ/02 prevent non-panic DKIM error from being sent to paniclog
3385 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
3388 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
3389 Notification from Dr Andrew Aitchison.
3391 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
3392 ExtendedDetectionInfo response format.
3393 Notification from John Horne.
3395 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
3398 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
3399 XSL and documented dependency on system catalogs, with examples of how
3402 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
3405 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
3406 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
3407 configuration file which is writeable by the Exim user or group.
3409 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
3410 of configuration files to cover files specified with the -C option if
3411 they are going to be used with root privileges, not just the default
3414 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
3415 option (effectively making it always true).
3417 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
3418 files to be used while preserving root privileges.
3420 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
3421 that rogue child processes cannot use them.
3423 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
3424 run-time user, instead of root.
3426 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
3427 Exim run-time user without dropping privileges.
3429 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
3430 result string, instead of calling string_vformat() twice with the same
3433 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
3434 for other users. Others should always drop root privileges if they use
3435 -C on the command line, even for a whitelisted configure file.
3437 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
3439 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
3445 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
3446 $data_path, and $header_path variables; fixed documentation bugs and
3449 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
3450 exipick to access non-standard spools, including the "frozen" queue
3453 NM/01 Bugzilla 965: Support mysql stored procedures.
3454 Patch from Alain Williams
3456 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
3458 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
3459 Patch from Andreas Metzler
3461 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
3462 Patch from Kirill Miazine
3464 NM/05 Bugzilla 671: Added umask to procmail example.
3466 JJ/03 installed exipick 20100323.0, fixing doc bug
3468 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
3469 directory. Notification and patch from Dan Rosenberg.
3471 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
3473 TK/02 Improve log output when DKIM signing operation fails.
3475 MH/01 Treat the transport option dkim_domain as a colon separated
3476 list, not as a single string, and sign the message with each element,
3477 omitting multiple occurences of the same signer.
3479 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
3480 Bugzilla 985, 986. Patch by Simon Arlott
3482 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
3483 Patch by Simon Arlott
3485 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
3486 MBX locking. Notification from Dan Rosenberg.
3492 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
3494 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
3496 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
3498 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
3500 NM/04 Bugzilla 915: Flag broken perl installation during build.
3506 TK/01 Added patch by Johannes Berg that expands the main option
3507 "spamd_address" if it starts with a dollar sign.
3509 TK/02 Write list of recipients to X-Envelope-Sender header when building
3510 the mbox-format spool file for content scanning (suggested by Jakob
3513 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
3514 (http://www.dcc-servers.net/) support via dccifd. Activated by
3515 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
3517 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
3518 by Mark Daniel Reidel <mr@df.eu>.
3520 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
3521 When building exim an external PCRE library is now needed -
3522 PCRE is a system library on the majority of modern systems.
3523 See entry on PCRE_LIBS in EDITME file.
3525 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
3526 conversation. Added nologin parameter to request.
3527 Patch contributed by Kirill Miazine.
3529 TF/01 Do not log submission mode rewrites if they do not change the address.
3531 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
3533 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
3534 log files in place. Contributed by Roberto Lima.
3536 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
3538 TF/03 Bugzilla 615: When checking the local_parts router precondition
3539 after a local_part_suffix or local_part_prefix option, Exim now
3540 does not use the address's named list lookup cache, since this
3541 contains cached lookups for the whole local part.
3543 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
3544 Robert Millan. Documentation is in experimental-spec.txt.
3546 TF/04 Bugzilla 668: Fix parallel build (make -j).
3548 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
3550 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
3551 Patch provided by Jan Srzednicki.
3553 TF/05 Leading white space used to be stripped from $spam_report which
3554 wrecked the formatting. Now it is preserved.
3556 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
3557 that they are available at delivery time.
3559 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
3561 TF/08 TLS error reporting now respects the incoming_interface and
3562 incoming_port log selectors.
3564 TF/09 Produce a more useful error message if an SMTP transport's hosts
3565 setting expands to an empty string.
3567 NM/06 Bugzilla 744: EXPN did not work under TLS.
3568 Patch provided by Phil Pennock.
3570 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
3571 Patch provided by Richard Godbee.
3573 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
3574 acl_smtp_notquit, added index entry.
3576 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
3577 Patch provided by Eugene Bujak.
3579 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
3580 accept(). Patch provided by Maxim Dounin.
3582 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
3583 Patch provided by Phil Pennock.
3585 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
3587 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
3588 Patch provided by Brad "anomie" Jorsch.
3590 NM/14 Bugzilla 622: Added timeout setting to transport filter.
3591 Patch provided by Dean Brooks.
3593 TK/05 Add native DKIM support (does not depend on external libraries).
3595 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
3596 Patch provided by Graeme Fowler.
3598 NM/16 Bugzilla 851: Documentation example syntax fix.
3600 NM/17 Changed NOTICE file to remove references to embedded PCRE.
3602 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
3605 NM/19 Bugzilla 745: TLS version reporting.
3606 Patch provided by Phil Pennock.
3608 NM/20 Bugzilla 167: bool: condition support.
3609 Patch provided by Phil Pennock.
3611 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
3612 clients. Patch provided by Phil Pennock.
3614 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
3615 Patch provided by Brad "anomie" Jorsch.
3617 NM/23 Bugzilla 687: Fix misparses in eximstats.
3618 Patch provided by Heiko Schlittermann.
3620 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
3621 Patch provided by Heiko Schlittermann.
3623 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
3624 plus update to original patch.
3626 NM/26 Bugzilla 799: Documentation correction for ratelimit.
3628 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
3629 Patch provided by David Brownlee.
3631 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
3633 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
3635 NM/30 Bugzilla 888: TLS documentation bugfixes.
3637 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
3639 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
3640 Unlike the original bugzilla I have changed all shell scripts in src tree.
3642 NM/33 Bugzilla 898: Transport filter timeout fix.
3643 Patch by Todd Rinaldo.
3645 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
3646 Patch by Serge Demonchaux.
3648 NM/35 Bugzilla 39: Base64 decode bug fixes.
3649 Patch by Jakob Hirsch.
3651 NM/36 Bugzilla 909: Correct connect() call in dcc code.
3653 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
3655 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
3657 NM/39 Bugzilla 911: Fixed MakeLinks build script.
3663 TK/01 Add preliminary DKIM support. Currently requires a forked version of
3664 ALT-N's libdkim that I have put here:
3665 http://duncanthrax.net/exim-experimental/
3667 Note to Michael Haardt: I had to rename some vars in sieve.c. They
3668 were called 'true' and it seems that C99 defines that as a reserved
3669 keyword to be used with 'bool' variable types. That means you could
3670 not include C99-style headers which use bools without triggering
3671 build errors in sieve.c.
3673 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
3674 as mailq or other aliases. Changed the --help handling significantly
3675 to do whats expected. exim_usage() emits usage/help information.
3677 SC/01 Added the -bylocaldomain option to eximstats.
3679 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
3681 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
3683 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
3689 PH/01 Another patch from the Sieve maintainer.
3691 PH/02 When an IPv6 address is converted to a string for single-key lookup
3692 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
3693 dots are used instead of colons so that keys in lsearch files need not
3694 contain colons. This was done some time before quoting was made available
3695 in lsearch files. However, iplsearch files do require colons in IPv6 keys
3696 (notated using the quote facility) so as to distinguish them from IPv4
3697 keys. This meant that lookups for IP addresses in host lists did not work
3698 for iplsearch lookups.
3700 This has been fixed by arranging for IPv6 addresses to be expressed with
3701 colons if the lookup type is iplsearch. This is not incompatible, because
3702 previously such lookups could never work.
3704 The situation is now rather anomalous, since one *can* have colons in
3705 ordinary lsearch keys. However, making the change in all cases is
3706 incompatible and would probably break a number of configurations.
3708 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
3711 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
3712 conversion specification without a maximum field width, thereby enabling
3713 a rogue spamd server to cause a buffer overflow. While nobody in their
3714 right mind would setup Exim to query an untrusted spamd server, an
3715 attacker that gains access to a server running spamd could potentially
3716 exploit this vulnerability to run arbitrary code as the Exim user.
3718 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
3719 $primary_hostname instead of what libspf2 thinks the hosts name is.
3721 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
3722 a directory entry by the name of the lookup key. Previously, if a
3723 symlink pointed to a non-existing file or a file in a directory that
3724 Exim lacked permissions to read, a lookup for a key matching that
3725 symlink would fail. Now it is enough that a matching directory entry
3726 exists, symlink or not. (Bugzilla 503.)
3728 PH/03 The body_linecount and body_zerocount variables are now exported in the
3731 PH/04 Added the $dnslist_matched variable.
3733 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
3734 This means they are set thereafter only if the connection becomes
3737 PH/06 Added the client_condition to authenticators so that some can be skipped
3738 by clients under certain conditions.
3740 PH/07 The error message for a badly-placed control=no_multiline_responses left
3741 "_responses" off the end of the name.
3743 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
3745 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
3746 (without spaces) instead of just copying the configuration text.
3748 PH/10 Added the /noupdate option to the ratelimit ACL condition.
3750 PH/11 Added $max_received_linelength.
3752 PH/12 Added +ignore_defer and +include_defer to host lists.
3754 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
3755 way in which PCRE > 7.0 is built.
3757 PH/14 Implemented queue_only_load_latch.
3759 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
3760 MAIL command. The effect was to mangle the value on 64-bit systems.
3762 PH/16 Another patch from the Sieve maintainer.
3764 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
3766 PH/18 If a system quota error occurred while trying to create the file for
3767 a maildir delivery, the message "Mailbox is full" was not appended to the
3768 bounce if the delivery eventually timed out. Change 4.67/27 below applied
3769 only to a quota excession during the actual writing of the file.
3771 PH/19 It seems that peer DN values may contain newlines (and other non-printing
3772 characters?) which causes problems in log lines. The DN values are now
3773 passed through string_printing() before being added to log lines.
3775 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
3776 and InterBase are left for another time.)
3778 PH/21 Added message_body_newlines option.
3780 PH/22 Guard against possible overflow in moan_check_errorcopy().
3782 PH/23 POSIX allows open() to be a macro; guard against that.
3784 PH/24 If the recipient of an error message contained an @ in the local part
3785 (suitably quoted, of course), incorrect values were put in $domain and
3786 $local_part during the evaluation of errors_copy.
3792 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
3793 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
3796 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
3797 issue a MAIL command.
3799 PH/02 In an ACL statement such as
3801 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
3803 if a client was not listed at all, or was listed with a value other than
3804 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
3805 the condition was not true (as it should be), so access was not denied.
3806 The bug was that the ! inversion was incorrectly passed on to the second
3807 item. This has been fixed.
3809 PH/03 Added additional dnslists conditions == and =& which are different from
3810 = and & when the dns lookup returns more than one IP address.
3812 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
3813 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
3815 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
3816 FSYNC, which compiles an option called disable_fsync that allows for
3817 bypassing fsync(). The documentation is heavily laced with warnings.
3819 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
3821 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
3822 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
3823 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
3824 including adding "make clean"; (3) Added -fPIC when compiling the test
3825 dynamically loaded module, to get rid of a warning.
3827 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
3828 message fails, move_frozen_messages = true and ignore_bounce_errors_after
3829 = 0s. The bug is otherwise harmless.
3831 PH/07 There was a bug in the dovecot authenticator such that the value of
3832 $auth1 could be overwritten, and so not correctly preserved, after a
3833 successful authentication. This usually meant that the value preserved by
3834 the server_setid option was incorrect.
3836 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
3838 PH/09 Installed PCRE release 7.0.
3840 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
3841 run for batched SMTP input. It is now run at the start of every message
3842 in the batch. While fixing this I discovered that the process information
3843 (output by running exiwhat) was not always getting set for -bs and -bS
3844 input. This is fixed, and it now also says "batched" for BSMTP.
3846 PH/11 Added control=no_pipelining.
3848 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
3849 patch, slightly modified), and move the expansion of helo_data till after
3850 the connection is made in the smtp transport (so it can use these
3853 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
3855 PH/14 Added log_selector = +pid.
3857 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
3859 PH/16 Add ${if forany and ${if forall.
3861 PH/17 Added dsn_from option to vary the From: line in DSNs.
3863 PH/18 Flush SMTP output before performing a callout, unless control =
3864 no_callout_flush is set.
3866 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
3867 was true (the default) a successful delivery failed to delete the retry
3868 item, thus causing premature timeout of the address. The bug is now
3871 PH/20 Added hosts_avoid_pipelining to the smtp transport.
3873 PH/21 Long custom messages for fakedefer and fakereject are now split up
3874 into multiline responses in the same way that messages for "deny" and
3875 other ACL rejections are.
3877 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
3878 with slight modification.
3880 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
3881 draft, changing the syntax and factoring some duplicate code".
3883 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
3884 for deliveries of the second and subsequent messages over the same SMTP
3887 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
3888 ${reduce, with only minor "tidies".
3890 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
3892 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
3893 expansion side effects.
3895 PH/27 When a message times out after an over-quota error from an Exim-imposed
3896 quota, the bounce message says "mailbox is full". This message was not
3897 being given when it was a system quota that was exceeded. It now should
3900 MH/03 Made $recipients available in local_scan(). local_scan() already has
3901 better access to the recipient list through recipients_list[], but
3902 $recipients can be useful in postmaster-provided expansion strings.
3904 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
3905 in the case of a MAIL command with additional options following the
3906 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
3907 were accidentally chopped off.
3909 PH/29 SMTP synchronization checks are implemented when a command is read -
3910 there is a check that no more input is waiting when there shouldn't be
3911 any. However, for some commands, a delay in an ACL can mean that it is
3912 some time before the response is written. In this time, more input might
3913 arrive, invalidly. So now there are extra checks after an ACL has run for
3914 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
3915 pipelining has not been advertised.
3917 PH/30 MH's patch to allow iscntrl() characters to be list separators.
3919 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
3920 returned in the SMTP response when smtp_return_error_details was false.
3921 This has been fixed.
3923 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
3924 instead of the C I/O that was originally supplied, because problems were
3925 reported on Solaris.
3927 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
3928 Exim which did not show up earlier: it was assuming that a call to
3929 SSL_CTX_set_info_callback() might give an error value. In fact, there is
3930 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
3931 was a macro that became an assignment, so it seemed to work. This has
3932 changed to a proper function call with a void return, hence the compile
3933 error. Exim's code has been fixed.
3935 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
3938 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
3940 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
3942 PH/37 If a message is not accepted after it has had an id assigned (e.g.
3943 because it turns out to be too big or there is a timeout) there is no
3944 "Completed" line in the log. When some messages of this type were
3945 selected by exigrep, they were listed as "not completed". Others were
3946 picked up by some special patterns. I have improved the selection
3947 criteria to be more general.
3949 PH/38 The host_find_failed option in the manualroute router can now be set
3950 to "ignore", to completely ignore a host whose IP address cannot be
3951 found. If all hosts are ignored, the behaviour is controlled by the new
3952 host_all_ignored option.
3954 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
3955 homing or because of multiple MX records with /mx) generated more than
3956 one IP address, and the following item turned out to be the local host,
3957 all the secondary addresses of the first item were incorrectly removed
3958 from the list, along with the local host and any following hosts (which
3959 is what is supposed to happen).
3961 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
3962 whoever called Exim into the -H file. In the case of the daemon it was
3963 behaving confusingly. When first started, it used values for whoever
3964 started the daemon, but after a SIGHUP it used the Exim user (because it
3965 calls itself on a restart). I have changed the code so that it now always
3968 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
3969 message are rejected with the same error (e.g. no authentication or bad
3970 sender address), and a DATA command is nevertheless sent (as can happen
3971 with PIPELINING or a stupid MUA), the error message that was given to the
3972 RCPT commands is included in the rejection of the DATA command. This is
3973 intended to be helpful for MUAs that show only the final error to their
3976 PH/42 Another patch from the Sieve maintainer.
3978 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
3979 Eximstats - Fixed some broken HTML links and added missing column headers
3981 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
3982 columns for Rejects, Temp Rejects, Ham, and Spam rows.
3984 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
3986 PH/43 Yet another patch from the Sieve maintainer.
3988 PH/44 I found a way to check for a TCP/IP connection going away before sending
3989 the response to the final '.' that terminates a message, but only in the
3990 case where the client has not sent further data following the '.'
3991 (unfortunately, this is allowed). However, in many cases there won't be
3992 any further data because there won't be any more messages to send. A call
3993 to select() can be used: if it shows that the input is "ready", there is
3994 either input waiting, or the socket has been closed. An attempt to read
3995 the next input character can distinguish the two cases. Previously, Exim
3996 would have sent an OK response which the client would never have see.
3997 This could lead to message repetition. This fix should cure that, at
3998 least in a lot of common cases.
4000 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
4001 advertised in response to EHLO.
4007 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
4008 fixed by 4.65/MH/01 (is this a record?) are fixed:
4010 (i) An empty string was always treated as zero by the numeric comparison
4011 operators. This behaviour has been restored.
4013 (ii) It is documented that the numeric comparison operators always treat
4014 their arguments as decimal numbers. This was broken in that numbers
4015 starting with 0 were being interpreted as octal.
4017 While fixing these problems I realized that there was another issue that
4018 hadn't been noticed. Values of message_size_limit (both the global option
4019 and the transport option) were treated as octal if they started with 0.
4020 The documentation was vague. These values are now always treated as
4021 decimal, and I will make that clear in the documentation.
4027 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
4028 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
4031 MH/01 Don't check that the operands of numeric comparison operators are
4032 integers when their expansion is in "skipping" mode (fixes bug
4033 introduced by 4.64-PH/07).
4035 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
4036 child addresses, Exim now panics and dies. Previously, because the count
4037 is held in a short int, deliveries were likely to be lost. As such a
4038 large number of recipients for a single message is ridiculous
4039 (performance will be very, very poor), I have chosen to impose a limit
4040 rather than extend the field.
4046 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
4047 leftover -K file (the existence of which was triggered by #402).
4048 While we were at it, introduced process PID as part of the -K
4049 filename. This should rule out race conditions when creating
4052 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
4053 processing considerably. Previous code took too long for large mails,
4054 triggering a timeout which in turn triggers #401.
4056 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
4057 in the DK code in transports.c. sendfile() is not really portable,
4058 hence the _LINUX specificness.
4060 TF/01 In the add_headers option to the mail command in an Exim filter,
4061 there was a bug that Exim would claim a syntax error in any
4062 header after the first one which had an odd number of characters
4065 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
4066 callout verification, Exim cached a "reject" for the entire domain. This
4067 is correct for most verifications, but it is not correct for a recipient
4068 verification with use_sender or use_postmaster set, because in that case
4069 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
4070 case of MAIL FROM:<> rejection from other early rejections (e.g.
4071 rejection of HELO). When verifying a recipient using a non-null MAIL
4072 address, the cache is ignored if it shows MAIL FROM:<> rejection.
4073 Whatever the result of the callout, the value of the domain cache is
4074 left unchanged (for any other kind of callout, getting as far as trying
4075 RCPT means that the domain itself is ok).
4077 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
4080 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
4081 manifest itself as EPIPE rather than ECONNECT. When tidying away a
4082 session, the daemon ignores ECONNECT errors and logs others; it now
4083 ignores EPIPE as well.
4085 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
4086 (quoted-printable decoding).
4088 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
4089 later the small subsequent patch to fix an introduced bug.
4091 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
4093 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
4095 PH/08 An error is now given if message_size_limit is specified negative.
4097 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
4098 to be given (somewhat) arbitrary names.
4100 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
4103 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
4104 miscellaneous code fixes
4106 PH/10 Added the log_reject_target ACL modifier to specify where to log
4109 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
4110 hostname. This is wrong, because it relates to the incoming message (and
4111 probably the interface on which it is arriving) and not to the outgoing
4112 callout (which could be using a different interface). This has been
4113 changed to use the value of the helo_data option from the smtp transport
4114 instead - this is what is used when a message is actually being sent. If
4115 there is no remote transport (possible with a router that sets up host
4116 addresses), $smtp_active_hostname is used.
4118 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
4119 tweaks were necessary in order to get it to work (see also 21 below):
4120 (a) The code assumed that strncpy() returns a negative number on buffer
4121 overflow, which isn't the case. Replaced with Exim's string_format()
4123 (b) There were several signed/unsigned issues. I just did the minimum
4124 hacking in of casts. There is scope for a larger refactoring.
4125 (c) The code used strcasecmp() which is not a standard C function.
4126 Replaced with Exim's strcmpic() function.
4127 (d) The code set only $1; it now sets $auth1 as well.
4128 (e) A simple test gave the error "authentication client didn't specify
4129 service in request". It would seem that Dovecot has changed its
4130 interface. Fortunately there's a specification; I followed it and
4131 changed what the client sends and it appears to be working now.
4133 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
4136 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
4137 address A is aliased to B and C, where B exists and C does not. Without
4138 -v the output is "A verified" because verification stops after a
4139 successful redirection if more than one address is generated. However,
4140 with -v the child addresses are also verified. Exim was outputting "A
4141 failed to verify" and then showing the successful verification for C,
4142 with its parentage. It now outputs "B failed to verify", showing B's
4143 parentage before showing the successful verification of C.
4145 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
4146 look up a TXT record in a specific list after matching in a combined
4149 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
4150 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
4151 they consult the DNS. I had assumed they would set it the way they
4152 wanted; and indeed my experiments on Linux seem to show that in some
4153 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
4154 To be on the safe side, however, I have now made the interface to
4155 host_find_byname() similar to host_find_bydns(), with an argument
4156 containing the DNS resolver options. The host_find_byname() function now
4157 sets these options at its start, just as host_find_bydns() does. The smtp
4158 transport options dns_qualify_single and dns_search_parents are passed to
4159 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
4160 of host_find_byname() use the default settings of RES_DEFNAMES
4161 (qualify_single) but not RES_DNSRCH (search_parents).
4163 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
4164 spool_read_header() do less string testing, by means of a preliminary
4165 switch on the second character of optional "-foo" lines. (This is
4166 overdue, caused by the large number of possibilities that now exist.
4167 Originally there were few.) While I was there, I also converted the
4168 str(n)cmp tests so they don't re-test the leading "-" and the first
4169 character, in the hope this might squeeze out yet more improvement.
4171 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
4172 flag allowing group syntax was set by the header_syntax check but not
4173 turned off, possible causing trouble later; (2) The flag was not being
4174 set at all for the header_verify test, causing "group"-style headers to
4175 be rejected. I have now set it in this case, and also caused header_
4176 verify to ignore an empty address taken from a group. While doing this, I
4177 came across some other cases where the code for allowing group syntax
4178 while scanning a header line wasn't quite right (mostly, not resetting
4179 the flag correctly in the right place). These bugs could have caused
4180 trouble for malformed header lines. I hope it is now all correct.
4182 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
4183 with the "reply" argument non-NULL. The code, however (which originally
4184 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
4185 but it didn't always do it. This confused somebody who was copying the
4186 code for some other use. I have removed all the tests.
4188 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
4189 feature that was used to support insecure browsers during the U.S. crypto
4190 embargo. It requires special client support, and Exim is probably the
4191 only MTA that supported it -- and would never use it because real RSA is
4192 always available. This code has been removed, because it had the bad
4193 effect of slowing Exim down by computing (never used) parameters for the
4194 RSA_EXPORT functionality.
4196 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
4197 authenticator to fail if there's a tab character in the incoming data
4198 (there should never be unless someone is messing about, as it's supposed
4199 to be base64-encoded). Also added, on Timo's advice, the "secured" option
4200 if the connection is using TLS or if the remote IP is the same as the
4201 local IP, and the "valid-client-cert option" if a client certificate has
4204 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
4205 authenticators. This can be used for authorization after authentication
4206 succeeds. (In the case of plaintext, it servers for both authentication
4209 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
4210 if any retry times were supplied.
4212 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
4213 connection, where there is no client IP address to check. In this
4214 situation, the verify now always succeeds.
4216 PH/25 Applied John Jetmore's -Mset patch.
4218 PH/26 Added -bem to be like -Mset, but loading a message from a file.
4220 PH/27 In a string expansion for a processed (not raw) header when multiple
4221 headers of the same name were present, leading whitespace was being
4222 removed from all of them, but trailing whitespace was being removed only
4223 from the last one. Now trailing whitespace is removed from each header
4224 before concatenation. Completely empty headers in a concatenation (as
4225 before) are ignored.
4227 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
4228 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
4230 PH/29 [Removed. This was a change that I later backed out, and forgot to
4231 correct the ChangeLog entry (that I had efficiently created) before
4232 committing the later change.]
4234 PH/30 Exim was sometimes attempting to deliver messages that had suffered
4235 address errors (4xx response to RCPT) over the same connection as other
4236 messages routed to the same hosts. Such deliveries are always "forced",
4237 so retry times are not inspected. This resulted in far too many retries
4238 for the affected addresses. The effect occurred only when there were more
4239 hosts than the hosts_max_try setting in the smtp transport when it had
4240 the 4xx errors. Those hosts that it had tried were not added to the list
4241 of hosts for which the message was waiting, so if all were tried, there
4242 was no problem. Two fixes have been applied:
4244 (i) If there are any address or message errors in an SMTP delivery, none
4245 of the hosts (tried or untried) are now added to the list of hosts
4246 for which the message is waiting, so the message should not be a
4247 candidate for sending over the same connection that was used for a
4248 successful delivery of some other message. This seems entirely
4249 reasonable: after all the message is NOT "waiting for some host".
4250 This is so "obvious" that I'm not sure why it wasn't done
4251 previously. Hope I haven't missed anything, but it can't do any
4252 harm, as the worst effect is to miss an optimization.
4254 (ii) If, despite (i), such a delivery is accidentally attempted, the
4255 routing retry time is respected, so at least it doesn't keep
4256 hammering the server.
4258 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
4259 in ${readsocket because some servers need this prod.
4261 PH/32 Added some extra debug output when updating a wait-xxx database.
4263 PH/33 The hint "could be header name not terminated by colon", which has been
4264 given for certain expansion errors for a long time, was not being given
4265 for the ${if def:h_colon_omitted{... case.
4267 PH/34 The spec says: "With one important exception, whenever a domain list is
4268 being scanned, $domain contains the subject domain." There was at least
4269 one case where this was not true.
4271 PH/35 The error "getsockname() failed: connection reset by peer" was being
4272 written to the panic log as well as the main log, but it isn't really
4273 panic-worthy as it just means the connection died rather early on. I have
4274 removed the panic log writing for the ECONNRESET error when getsockname()
4277 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
4278 runs only) independently of the message's sender address. This meant
4279 that, if the 4xx error was in fact related to the sender, a different
4280 message to the same recipient with a different sender could confuse
4281 things. In particular, this can happen when sending to a greylisting
4282 server, but other circumstances could also provoke similar problems.
4283 I have changed the default so that the retry time for these errors is now
4284 based a combination of the sender and recipient addresses. This change
4285 can be overridden by setting address_retry_include_sender=false in the
4288 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
4289 remote server are returned as part of bounce messages. This was not
4290 happening for LMTP over a pipe (the lmtp transport), but now it is the
4291 same for both kinds of LMTP.
4293 PH/38 Despite being documented as not happening, Exim was rewriting addresses
4294 in header lines that were in fact CNAMEs. This is no longer the case.
4296 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
4297 and queue runs started by the daemon processed all messages. This has
4298 been fixed so that -R and -S can now usefully be given with -q<time>.
4300 PH/40 Import PCRE release 6.7 (fixes some bugs).
4302 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
4304 PH/42 Give an error if -q is specified more than once.
4306 PH/43 Renamed the variables $interface_address and $interface_port as
4307 $received_ip_address and $received_port, to make it clear that these
4308 values apply to message reception, and not to the outgoing interface when
4309 a message is delivered. (The old names remain recognized, of course.)
4311 PH/44 There was no timeout on the connect() call when using a Unix domain
4312 socket in the ${readsocket expansion. There now is.
4314 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
4315 be meaningful with "accept".
4317 SC/01 Eximstats V1.43
4318 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
4320 SC/02 Eximstats V1.44
4321 Use a glob alias rather than an array ref in the generated
4322 parser. This improves both readability and performance.
4324 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
4325 Collect SpamAssassin and rejection statistics.
4326 Don't display local sender or destination tables unless
4327 there is data to show.
4328 Added average volumes into the top table text output.
4330 SC/04 Eximstats V1.46
4331 Collect data on the number of addresses (recipients)
4332 as well as the number of messages.
4334 SC/05 Eximstats V1.47
4335 Added 'Message too big' to the list of mail rejection
4336 reasons (thanks to Marco Gaiarin).
4338 SC/06 Eximstats V1.48
4339 Mainlog lines which have GMT offsets and are too short to
4340 have a flag are now skipped.
4342 SC/07 Eximstats V1.49 (Alain Williams)
4343 Added the -emptyok flag.
4345 SC/08 Eximstats V1.50
4346 Fixes for obtaining the IP address from reject messages.
4348 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
4349 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
4350 whitespace changes from 4.64-PH/27
4352 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
4355 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
4356 are found, allow negative numbers in numeric criteria)
4358 JJ/06 exipick.20061117.2, added new $message_body_missing variable
4360 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
4361 to match changes made in 4.64-PH/43
4363 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
4365 PH/47 Put in an explicit test for a DNS lookup of an address record where the
4366 "domain" is actually an IP address, and force a failure. This locks out
4367 those revolvers/nameservers that support "A-for-A" lookups, in
4368 contravention of the specifications.
4370 PH/48 When a host name was looked up from an IP address, and the subsequent
4371 forward lookup of the name timed out, the host name was left in
4372 $sender_host_name, contrary to the specification.
4374 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
4375 restricted to single-key lookups, Exim was not diagnosing an error if
4376 * or *@ was used with a query-style lookup.
4378 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
4380 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
4381 long ago, but noone interested enough thought of it. Let's just say that
4382 the "1.1" means that there are some new functions that weren't there at
4383 some point in the past.
4385 PH/51 Error processing for expansion failure of helo_data from an smtp
4386 transport during callout processing was broken.
4388 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
4389 tested/used via the -bh/-bhc/-bs options.
4391 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
4392 bug, fixed in subsequent PCRE releases).
4394 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
4395 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
4397 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
4403 SC/01 Use a glob alias rather than an array ref in eximstats generated
4404 parser. This improves both readability and performance.
4406 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
4407 Don't display local sender or destination tables in eximstats unless
4408 there is data to show.
4409 Added average volumes into the eximstats top table text output.
4411 SC/03 Collect data on the number of addresses (recipients) as well
4412 as the number of messages in eximstats.
4414 TF/01 Correct an error in the documentation for the redirect router. Exim
4415 does (usually) call initgroups() when daemonizing.
4417 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
4418 with consistent privilege compared to when running as a daemon.
4420 TF/03 Note in the spec that $authenticated_id is not set for local
4421 submissions from trusted users.
4423 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
4424 Thanks to Dean Brooks <dean@iglou.com> for the patch.
4426 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
4427 by adding some example configuration directives to the default
4428 configuration file. A little bit of work is required to uncomment the
4429 directives and define how usernames and passwords are checked, but
4430 there is now a framework to start from.
4432 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
4433 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
4434 without this. I don't know how relevant this is to other LDAP libraries.
4436 PH/02 Add the verb name to the "unknown ACL verb" error.
4438 PH/03 Magnus Holmgren's patch for filter_prepend_home.
4440 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
4442 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
4443 directory not expanded when it should be if an expanded home directory
4444 was set for the address (which is overridden by the transport).
4446 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
4449 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
4450 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
4451 because it is too late at that time, and has no effect.
4453 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
4454 security issue with \' (bugzilla #107). I could not use the
4455 PQescapeStringConn() function, because it needs a PGconn value as one of
4458 PH/08 When testing addresses using -bt, indicate those final addresses that
4459 are duplicates that would not cause an additional delivery. At least one
4460 person was confused, thinking that -bt output corresponded to deliveries.
4461 (Suppressing duplicates isn't a good idea as you lose the information
4462 about possibly different redirections that led to the duplicates.)
4464 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
4465 systems where poll() doesn't work, in particular OS X.
4467 PH/10 Added more information to debugging output for retry time not reached.
4469 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
4470 operations in malware.c.
4472 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
4475 PH/13 If write_rejectlog was set false when logging was sent to syslog with
4476 syslog_duplication set false, log lines that would normally be written
4477 both the the main log and to the reject log were not written to syslog at
4480 PH/14 In the default configuration, change the use of "message" in ACL warn
4481 statements to "add_header".
4483 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
4484 not followed by a command (e.g. "seen endif").
4486 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
4487 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
4490 PH/17 Added extra conditions to the default value of delay_warning_condition
4494 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
4495 { match{$h_precedence:}{(?i)bulk|list|junk} } \
4496 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
4499 The Auto-Submitted: and various List- headers are standardised, whereas I
4500 don't think Precedence: ever was.
4502 PH/18 Refactored debugging code in route_finduser() to show more information,
4503 in particular, the error code if getpwnam() issues one.
4505 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
4506 This is apparently needed in addition to the PH/07 change above to avoid
4507 any possible encoding problems.
4509 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
4510 but not after initializing Perl.
4512 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
4513 output them only if debugging. By default they are written stderr,
4514 apparently, which is not desirable.
4516 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
4519 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
4522 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
4524 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
4525 authenticated or an ident call has been made. Suppress the default
4526 values for $authenticated_id and $authenticated_sender (but permit -oMai
4527 and -oMas) when testing with -bh.
4529 PH/24 Re-jigged the order of the tests in the default configuration so that the
4530 tests for valid domains and recipients precede the DNS black list and CSA
4531 tests, on the grounds that those ones are more expensive.
4533 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
4534 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
4535 This bug exists in every version of Exim that I still have, right back to
4538 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
4539 However, an attempt to turn on case-sensitivity in a regex key by
4540 including (?-i) didn't work because the subject string was already
4541 lowercased, and the effects were non-intuitive. It turns out that a
4542 one-line patch can be used to allow (?-i) to work as expected.
4548 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
4549 other effects) broke the use of negated acl sub-conditions.
4551 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
4554 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
4555 "Deny" causes Exim to reject the incoming connection with a 554 error.
4556 Unfortunately, if there is a major crisis, such as a disk failure,
4557 tcp-wrappers gives "deny", whereas what one would like would be some
4558 kind of temporary error. A kludge has been added to help with this.
4559 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
4560 554 error is used if errno is still zero or contains ENOENT (which occurs
4561 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
4564 PH/03 Add -lutil to the default FreeBSD LIBS setting.
4566 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
4567 errors. Otherwise a message that provokes a temporary error (when other
4568 messages do not) can cause a whole host to time out.
4570 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
4571 addresses were routed directly to files or pipes from a redirect router.
4572 File deliveries just didn't batch; pipe deliveries might have suffered
4575 PH/06 A failure to get a lock for a hints database would erroneously always say
4576 "Failed to get write lock", even when it was really a read lock.
4578 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
4579 of 0600. This has been changed to use the value of the lockfile_mode
4580 option (which defaults to 0600).
4582 PH/08 Applied small patch from the Sieve maintainer.
4584 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
4585 folder from quota calculations, a direct delivery into this folder messed
4586 up the contents of the maildirsize file. This was because the regex was
4587 used only to exclude .Trash (or whatever) when the size of the mailbox
4588 was calculated. There was no check that a delivery was happening into an
4589 excluded directory. This bug has been fixed by ignoring all quota
4590 processing for deliveries into excluded directories.
4592 PH/10 Added the maildirfolder_create_regex option to appendfile.
4598 PH/01 The code for finding all the local interface addresses on a FreeBSD
4599 system running IPv6 was broken. This may well have applied to all BSD
4600 systems, as well as to others that have similar system calls. The broken
4601 code found IPv4 interfaces correctly, but gave incorrect values for the
4602 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
4603 that it would not match correctly against @[] and not recognize the IPv6
4606 PH/02 The ipliteral router was not recognizing addresses of the form user@
4607 [ipv6:....] because it didn't know about the "ipv6:" prefix.
4609 PH/03 Added disable_ipv6.
4611 PH/04 Changed $reply_address to use the raw form of the headers instead of the
4612 decoded form, because it is most often used to construct To: headers
4613 lines in autoreplies, and the decoded form may well be syntactically
4614 invalid. However, $reply_address has leading white space removed, and all
4615 newlines turned into spaces so that the autoreply transport does not
4618 PH/05 If group was specified without a user on a router, and no group or user
4619 was specified on a transport, the group from the router was ignored.
4621 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
4622 for visible compile-time settings that can be used to change these
4623 numbers, for those that want even more. Backwards compatibility with old
4624 spool files has been maintained. However, going back to a previous Exim
4625 release will lost any variables that are in spool files.
4627 PH/07 Two small changes when running in the test harness: increase delay when
4628 passing a TCP/IP connection to a new process, in case the original
4629 process has to generate a bounce, and remove special handling of
4630 127.0.0.2 (sic), which is no longer necessary.
4632 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
4633 be the same on different OS.
4635 PH/09 Moved a debug statement in filter processing to avoid a race problem when
4638 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
4639 whether --show-vars was specified or not
4641 JJ/02 exipick: Added support for new ACL variable spool format introduced
4644 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
4645 syntactically invalid From: or Reply-to: line, and a filter used this to
4646 generate an autoreply, and therefore failed to obtain an address for the
4647 autoreply, Exim could try to deliver to a non-existent relative file
4648 name, causing unrelated and misleading errors. What now happens is that
4649 it logs this as a hard delivery error, but does not attempt to create a
4652 PH/11 The exinext utility has a -C option for testing purposes, but although
4653 the given file was scanned by exinext itself; it wasn't being passed on
4654 when Exim was called.
4656 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
4657 an end-of-file indication when reading a command response.
4659 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
4660 compiled. In many other places in Exim, IPv6 addresses are always
4661 recognized, so I have changed this. It also means that IPv4 domain
4662 literals of the form [IPV4:n.n.n.n] are now always recognized.
4664 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
4665 used if the router is not running as root, for example, when verifying at
4666 ACL time, or when using -bh. The debugging output from this situation was
4667 non-existent - all you got was a failure to exec. I have made two
4670 (a) Failures to set uid/gid, the current directory, or a process leader
4671 in a subprocess such as that created by queryprogram now generate
4672 suitable debugging output when -d is set.
4674 (b) The queryprogram router detects when it is not running as root,
4675 outputs suitable debugging information if -d is set, and then runs
4676 the subprocess without attempting to change uid/gid.
4678 PH/15 Minor change to Makefile for building test_host (undocumented testing
4681 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
4682 additional section of a DNS packet that returns MX or SRV records.
4683 Instead, it always explicitly searches for A/AAAA records. This avoids
4684 major problems that occur when a DNS server includes only records of one
4685 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
4686 fixed another bug: if SRV records were looked up and the corresponding
4687 address records were *not* found in the additional section, the port
4688 values from the SRV records were lost.
4690 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
4691 using the correct key (the original address) when searching the retry
4692 rules in order to find which one to use for generating the retry hint.
4694 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
4695 adding the default one. Similarly, if it contains a Reply-To: header, the
4696 errors_reply_to option, if set, is not used.
4698 PH/19 When calculating a retry time, Exim used to measure the "time since
4699 failure" by looking at the "first failed" field in the retry record. Now
4700 it does not use this if it is later than than the arrival time of the
4701 message. Instead it uses the arrival time. This makes for better
4702 behaviour in cases where some deliveries succeed, thus re-setting the
4703 "first failed" field. An example is a quota failure for a huge message
4704 when small messages continue to be delivered. Without this change, the
4705 "time since failure" will always be short, possible causing more frequent
4706 delivery attempts for the huge message than are intended.
4707 [Note: This change was subsequently modified - see PH/04 for 4.62.]
4709 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
4710 $1, $2, $3) because the numerical variables can be reset during some
4711 expansion items (e.g. "match"), thereby losing the authentication data.
4713 PH/21 Make -bV show the size of off_t variables so that the test suite can
4714 decide whether to run tests for quotas > 2G.
4716 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
4717 mailbox_size, and mailbox_filecount in the appendfile transport. If a
4718 filecount value is greater than 2G or if a quota value is greater than 2G
4719 on a system where the size of off_t is not greater than 4, a panic error
4722 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
4723 never match. The debug and -bh output now contains an explicit error
4724 message indicating a malformed IPv4 address or mask.
4726 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
4727 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
4728 PH/23 above applies.
4730 PH/25 Do not write to syslog when running in the test harness. The only
4731 occasion when this arises is a failure to open the main or panic logs
4732 (for which there is an explicit test).
4734 PH/26 Added the /no_tell option to "control=freeze".
4736 PH/27 If a host name lookup failed very early in a connection, for example, if
4737 the IP address matched host_lookup and the reverse lookup yielded a name
4738 that did not have a forward lookup, an error message of the form "no IP
4739 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
4740 could be logged. Now it outputs the IP address instead of "NULL".
4742 PH/28 An enabling patch from MH: add new function child_open_exim2() which
4743 allows the sender and the authenticated sender to be set when
4744 submitting a message from within Exim. Since child_open_exim() is
4745 documented for local_scan(), the new function should be too.
4747 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
4748 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
4749 results in an empty string is now treated as unset.
4751 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
4753 PH/31 Added sender_verify_fail logging option.
4755 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
4756 needed by RFC 822 but not by RFC 2822 was commented out. I have now
4757 tidied the source and removed it altogether.
4759 PH/33 When a queue run was abandoned because the load average was too high, a
4760 log line was always written; now it is written only if the queue_run log
4761 selector is set. In addition, the log line for abandonment now contains
4762 information about the queue run such as the pid. This is always present
4763 in "start" and "stop" lines but was omitted from the "abandon" line.
4765 PH/34 Omit spaces between a header name and the colon in the error message that
4766 is given when verify = headers_syntax fails (if there are lots of them,
4767 the message gets confusing).
4769 PH/35 Change the default for dns_check_names_pattern to allow slashes within
4770 names, as there are now some PTR records that contain slashes. This check
4771 is only to protect against broken name servers that fall over on strange
4772 characters, so the fact that it applies to all lookups doesn't matter.
4774 PH/36 Now that the new test suite is complete, we can remove some of the
4775 special code in Exim that was needed for the old test suite. For example,
4776 sorting DNS records because real resolvers return them in an arbitrary
4777 order. The new test suite's fake resolver always returns records in the
4780 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
4781 bounces) except when queue_only is set, to avoid logging races between
4782 the different processes.
4784 PH/38 Panic-die if .include specifies a non-absolute path.
4786 PH/39 A tweak to the "H" retry rule from its user.
4788 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
4789 a label. They prevented compilation on older perls.
4791 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
4792 a warning to be raised on newish perls.
4794 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
4795 on queue. Changes to match documented behaviour of showing count of
4796 messages matching specified criteria.
4798 PH/40 Changed the default ident timeout from 30s to 5s.
4800 PH/41 Added support for the use of login_cap features, on those BSD systems
4801 that have them, for controlling the resources used by pipe deliveries.
4803 PH/42 The content-scanning code uses fopen() to create files in which to put
4804 message data. Previously it was not paying any attention to the mode of
4805 the files. Exim runs with umask(0) because the rest of the code creates
4806 files with open(), and sets the required mode explicitly. Thus, these
4807 files were ending up world-writeable. This was not a big issue, because,
4808 being within the spool directory, they were not world-accessible. I have
4809 created a function called modefopen, which takes an additional mode
4810 argument. It sets umask(777), creates the file, chmods it to the required
4811 mode, then resets the umask. All the relevant calls to fopen() in the
4812 content scanning code have been changed to use this function.
4814 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
4815 to 24 hours. This avoids potential overflow problems when processing G
4816 and H retry rules. I suspect nobody ever tinkers with this value.
4818 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
4820 PH/45 When the plaintext authenticator is running as a client, the server's
4821 challenges are checked to ensure they are valid base64 strings. By
4822 default, the authentication attempt is cancelled if an invalid string is
4823 received. Setting client_ignore_invalid_base64 true ignores these errors.
4824 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
4825 they are received. Thus, the responses can be made to depend on the
4826 challenges. If an invalid string is ignored, an empty string is placed in
4829 PH/46 Messages that are created by the autoreply transport now contains a
4830 References: header, in accordance with RFCs 2822 and 3834.
4832 PH/47 Added authenticated_sender_force to the smtp transport.
4834 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
4836 PH/49 Installed latest patch from the Sieve maintainer.
4838 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
4839 was also set, the appendfile transport was unnecessarily scanning a
4840 directory of message files (e.g. for maildir delivery) to find the count
4841 of files (along with the size), even though it did not need this
4842 information. It now does the scan only if it needs to find either the
4843 size of the count of files.
4845 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
4847 PH/52 Two bugs concerned with error handling when the smtp transport is
4850 (i) Exim was not creating retry information for temporary errors given
4851 for individual recipients after the DATA command when the smtp transport
4852 was used in LMTP mode. This meant that they could be retried too
4853 frequently, and not timed out correctly.
4855 (ii) Exim was setting the flag that allows error details to be returned
4856 for LMTP errors on RCPT commands, but not for LMTP errors for individual
4857 recipients that were returned after the DATA command.
4859 PH/53 This is related to PH/52, but is more general: for any failing address,
4860 when detailed error information was permitted to be returned to the
4861 sender, but the error was temporary, then after the final timeout, only
4862 "retry timeout exceeded" was returned. Now it returns the full error as
4863 well as "retry timeout exceeded".
4865 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
4866 do this, and (what is worse) MTAs that accept it.
4868 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
4869 will now be deprecated.
4871 PH/56 New os.c-cygwin from the Cygwin maintainer.
4873 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
4874 formats (previously only available in exim formats via -bpr, -bpru,
4875 and -bpra. Now also available in native and exiqgrep formats)
4877 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
4878 with very large, slow to parse queues
4880 JJ/08 exipick: added ! as generic prefix to negate any criteria format
4882 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
4884 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
4885 responses to authentication challenges, though it was showing the
4886 challenges; (ii) I've removed the CR characters from the debug output for
4889 PH/58 Allow for the insertion of a newline as well as a space when a string
4890 is turned into more than one encoded-word during RFC 2047 encoding. The
4891 Sieve code now uses this.
4893 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
4894 data_4xx, lost_connection, tls_required.
4896 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
4897 message was being sent as an SMTP response.
4899 PH/61 Add -l and -k options to exicyclog.
4901 PH/62 When verifying, if an address was redirected to one new address, so that
4902 verification continued, and the new address failed or deferred after
4903 having set something in $address_data, the value of $address_data was not
4904 passed back to the ACL. This was different to the case when no
4905 redirection occurred. The value is now passed back in both cases.
4907 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
4908 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
4909 use login_cap.h, so on its own it isn't the distinguishing feature. The
4910 new name refers directly to the setclassresources() function.
4912 PH/65 Added configuration files for NetBSD3.
4914 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
4916 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
4917 is preferred over IPv4.
4919 PH/68 The bounce_return_message and bounce_return_body options were not being
4920 honoured for bounces generated during the reception of non-SMTP messages.
4921 In particular, this applied to messages rejected by the ACL. This bug has
4922 been fixed. However, if bounce_return_message is true and bounce_return_
4923 body is false, the headers that are returned for a non-SMTP message
4924 include only those that have been read before the error was detected.
4925 (In the case of an ACL rejection, they have all been read.)
4927 PH/69 The HTML version of the specification is now built in a directory called
4928 spec_html instead of spec.html, because the latter looks like a path with
4929 a MIME-type, and this confuses some software.
4931 PH/70 Catch two compiler warnings in sieve.c.
4933 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
4934 function verify_get_ident() calls ip_connect() to connect a socket, but
4935 if the "connect()" function timed out, ip_connect() used to close the
4936 socket. However, verify_get_ident() also closes the socket later, and in
4937 between Exim writes to the log, which may get opened at this point. When
4938 the socket was closed in ip_connect(), the log could get the same file
4939 descriptor number as the socket. This naturally causes chaos. The fix is
4940 not to close the socket in ip_connect(); the socket should be closed by
4941 the function that creates it. There was only one place in the code where
4942 this was missing, in the iplookup router, which I don't think anybody now
4943 uses, but I've fixed it anyway.
4945 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
4946 well as to direct DNS lookups. Otherwise the handling of names in host
4947 lists is inconsistent and therefore confusing.
4953 PH/01 Two changes to the default runtime configuration:
4955 (1) Move the checks for relay_from_hosts and authenticated clients from
4956 after to before the (commented out) DNS black list checks.
4958 (2) Add control=submission to the relay_from_hosts and authenticated
4959 clients checks, on the grounds that messages accepted by these
4960 statements are most likely to be submissions.
4962 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
4964 (1) Generate an error if the third argument for the ${prvs expansion is
4967 (2) Treat a missing third argument of ${prvscheck as if it were an empty
4970 (3) Reset the variables that are obtained from the first argument of
4971 ${prvscheck and used in the second argument before leaving the code,
4972 because their memory is reclaimed, so using them afterwards may do
4975 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
4976 one (it's much easier than Tom thought :-).
4978 (5) Because of (4), we can now allow for the use of $prvscheck_result
4979 inside the third argument.
4981 PH/03 For some reason, the default setting of PATH when running a command from
4982 a pipe transport was just "/usr/bin". I have changed it to
4985 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
4986 anything to be listed in the output from -bV.
4988 PH/05 When a filter generated an autoreply, the entire To: header line was
4989 quoted in the delivery log line, like this:
4991 => >A.N.Other <ano@some.domain> <original@ddress> ...
4993 This has been changed so that it extracts the operative address. There
4994 may be more than one such address. If so, they are comma-separated, like
4997 => >ano@some.domain,ona@other.domain <original@ddress> ...
4999 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
5000 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
5001 not being looked up in the rDNS to get a host name, Exim was showing the
5002 IP address twice in Received: lines, even though the IP addresses were
5003 identical. For example:
5005 Received: from [1.2.3.4] (helo=[1.2.3.4])
5007 However, if the real host name was known, it was omitting the HELO data
5008 if it matched the actual IP address. This has been tidied up so that it
5009 doesn't show the same IP address twice.
5011 PH/07 When both +timestamp and +memory debugging was on, the value given by
5012 $tod_xxx expansions could be wrong, because the tod_stamp() function was
5013 called by the debug printing, thereby overwriting the timestamp buffer.
5014 Debugging no longer uses the tod_stamp() function when +timestamp is set.
5016 PH/08 When the original message was included in an autoreply transport, it
5017 always said "this is a copy of the message, including all the headers",
5018 even if body_only or headers_only was set. It now gives an appropriate
5021 PH/09 Applied a patch from the Sieve maintainer which:
5023 o fixes some comments
5024 o adds the (disabled) notify extension core
5025 o adds some debug output for the result of if/elsif tests
5026 o points to the current vacation draft in the documentation
5027 and documents the missing references header update
5031 o fixes a bug in processing the envelope test (when testing
5032 multiple envelope elements, the last element determined the
5035 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
5036 Electronic Mail") by including:
5038 Auto-submitted: auto-generated
5040 in the messages that it generates (bounce messages and others, such as
5041 warnings). In the case of bounce messages for non-SMTP messages, there was
5042 also a typo: it was using "Auto_submitted" (underscore instead of
5043 hyphen). Since every message generated by Exim is necessarily in response
5044 to another message, thes have all been changed to:
5046 Auto-Submitted: auto-replied
5048 in accordance with these statements in the RFC:
5050 The auto-replied keyword:
5052 - SHOULD be used on messages sent in direct response to another
5053 message by an automatic process,
5055 - MUST NOT be used on manually-generated messages,
5057 - MAY be used on Delivery Status Notifications (DSNs) and Message
5058 Disposition Notifications (MDNs),
5060 - MUST NOT be used on messages generated by automatic or periodic
5061 processes, except for messages which are automatic responses to
5064 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
5065 to the default Received: header definition.
5067 PH/12 Added log selector acl_warn_skipped (default on).
5069 PH/13 After a successful wildlsearch lookup, discard the values of numeric
5070 variables because (a) they are in the wrong storage pool and (b) even if
5071 they were copied, it wouldn't work properly because of the caching.
5073 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
5074 checking when decoding. Apparently there are clients that generate
5075 overlong encoded strings. Why am I not surprised?
5077 PH/15 If the first argument of "${if match_address" was not empty, but did not
5078 contain an "@" character, Exim crashed. Now it writes a panic log message
5079 and treats the condition as false.
5081 PH/16 In autoreply, treat an empty string for "once" the same as unset.
5083 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
5084 extension "envelope-auth". The code is finished and in agreement with
5085 other implementations, but there is no documentation so far and in fact,
5086 nobody wrote the draft yet. This extension is currently #undef'ed, thus
5087 not changing the active code.
5089 Print executed "if" and "elsif" statements when debugging is used. This
5090 helps a great deal to understand what a filter does.
5092 Document more things not specified clearly in RFC3028. I had all this
5093 sorted out, when out of a sudden new issues came to my mind. Oops."
5095 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
5098 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
5099 canonical form (as documented). However, after a host name lookup from
5100 the IP address, check_host() was doing a simple string comparison with
5101 addresses acquired from the DNS when checking that the found name did
5102 have the original IP as one of its addresses. Since any found IPv6
5103 addresses are likely to be in abbreviated form, the comparison could
5104 fail. Luckily, there already exists a function for doing the comparison
5105 by converting both addresses to binary, so now that is used instead of
5106 the text comparison.
5108 PH/20 There was another similar case to PH/19, when a complete host name was
5109 given in a host list; looking up its IP address could give an abbreviated
5110 form, whereas the current host's name might or might not be abbreviated.
5111 The same fix has been applied.
5117 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
5118 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
5121 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
5122 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
5124 PH/03 Typo: missing ".o" in src/pcre/Makefile.
5126 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
5127 header line, restrict the check to what is listed in RFCs 2369 and 2929.
5128 Also, for "Auto-Submitted", treat anything other than "no" as
5129 non-personal, in accordance with RFC 3834. (Previously it treated
5130 anything starting "auto-" as non-personal.)
5132 TF/01 The control=submission/name=... option had a problem with syntax
5133 errors if the name included a slash character. The /name= option
5134 now slurps the rest of the string, so it can include any characters
5135 but it must come last in the list of options (after /sender_retain
5138 PH/05 Some modifications to the interface to the fake nameserver for the new
5146 TK/01 Added the "success_on_redirect" address verification option. See
5147 NewStuff for rationale and an example.
5149 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
5151 PH/02 Patch to exigrep to allow it to work on syslog lines.
5153 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
5154 fread() to skip over the body file's header line, because in Cygwin the
5155 header line is locked and is inaccessible.
5157 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
5158 co-exist for some time) to make it clear that it is the Exim ID that is
5159 referenced, not the Message-ID: header line.
5161 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
5162 string_format() function, because snprintf() does not exist on all
5165 PH/06 The use of forbid_filter_existstest now also locks out the use of the
5166 ${stat: expansion item.
5168 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
5169 protocol synchronization error", to keep the pedants happy.
5171 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
5172 well as for IRIX systems, when gcc is being used. See the host.c source
5175 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
5177 PH/10 Named domain lists were not working if used in a queue_smtp_domains
5180 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
5181 transport and to the smtp transport in LMTP mode.
5183 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
5185 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
5186 run a filter in a subprocess. This could lead to confusion in subsequent
5187 lookups in the parent process. There should also be a search_tidyup() at
5188 the end of the subprocess.
5190 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
5191 only if the host matched helo_try_verify_hosts, which caused the
5192 verification to occur when the EHLO/HELO command was issued. The ACL just
5193 tested the remembered result. Now, if a previous verification attempt has
5194 not happened, "verify = helo" does it there and then.
5196 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
5198 TK/03 Fix log output including CR from clamd.
5200 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
5201 exist provoked a memory error which could cause a segfault.
5203 PH/15 Installed PCRE 6.2
5205 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
5207 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
5208 of the problem. Specifically, suggested +O2 rather than +O1 for the
5211 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
5213 PH/20 If a delivery was routed to a non-standard port by means of an SRV
5214 record, the port was not correctly logged when the outgoing_port log
5215 selector was set (it logged the transort's default port).
5217 PH/21 Added support for host-specific ports to manualroute, queryprogram,
5218 fallback_hosts, and "hosts" in the smtp transport.
5220 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
5221 host errors such as "Connection refused".
5223 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
5224 authentication with radiusclient 0.4.9:
5226 - Error returned from rc_read_config was caught wrongly
5227 - Username/password not passed on to radius server due to wrong length.
5229 The presumption is that some radiusclient API changes for 4.51/PH/17
5230 were not taken care of correctly. The code is still untested by me (my
5231 Linux distribution still has 0.3.2 of radiusclient), but it was
5232 contributed by a Radius user.
5234 PH/24 When doing a callout, the value of $domain wasn't set correctly when
5235 expanding the "port" option of the smtp transport.
5237 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
5238 while reading a MIME header. Thanks to Tom Hughes for a patch.
5240 PH/24 Include config.h inside local_scan.h so that configuration settings are
5243 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
5244 that in an ACL for RCPT (for example), you can examine exactly what was
5247 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
5248 commands, but it was not correctly comparing the address with the actual
5249 client host address. Thus, it would show the EHLO address in Received:
5250 header lines when this was not necessary.
5252 PH/27 Added the % operator to ${eval:}.
5254 PH/28 Exim tries to create and chdir to its spool directory when it starts;
5255 it should be ignoring failures (because with -C, for example, it has lost
5256 privilege). It wasn't ignoring creation failures other than "already
5259 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
5262 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
5263 because an input line was too long, either on its own, or by virtue of
5264 too many continuations, the temporary file was not being removed, and the
5265 return code was incorrect.
5267 PH/31 Missing "BOOL" in function definition in filtertest.c.
5269 PH/32 Applied Sieve patches from the maintainer.
5271 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
5273 PH/33 Added "verify = not_blind".
5275 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
5276 Local/Makefile (with some defaults set). These are used in built scripts
5277 such as exicyclog, but they have never been used in the exim_install
5278 script (though there are many overriding facilities there). I have
5279 arranged that the exim_install script now takes note of these two
5282 PH/35 Installed configuration files for Dragonfly.
5284 PH/36 When a locally submitted message by a trusted user did not contain a
5285 From: header, and the sender address was obtained from -f or from an SMTP
5286 MAIL command, and the trusted user did not use -F to supply a sender
5287 name, $originator_name was incorrectly used when constructing a From:
5288 header. Furthermore, $originator_name was used for submission mode
5289 messages from external hosts without From: headers in a similar way,
5290 which is clearly wrong.
5292 PH/37 Added control=suppress_local_fixups.
5294 PH/38 When log_selector = +received_sender was set, and the addition of the
5295 sender made the log line's construction buffer exactly full, or one byte
5296 less than full, an overflow happened when the terminating "\n" was
5299 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
5300 when the result of a list match is failure because a DNS lookup failed.
5302 PH/40 RM_COMMAND is now used in the building process.
5304 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
5305 the "build-* directories that it finds.
5307 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
5308 address in a domain literal was a prefix of an interface address.
5310 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
5311 when verifying a sender address, unless rewrite_headers is false.
5313 PH/44 Wrote a long comment about why errors_to addresses are verified as
5314 recipients, not senders.
5316 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
5317 the ratelimit ACL was added.
5319 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
5321 PH/46 Added extra information about PostgreSQL errors to the error string.
5323 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
5324 avoiding the need to install special zones in a real server. This is
5325 backwards compatible; if it can't find the fake resolver, it drops back.
5326 Thus, both old and new test suites can be run.
5328 TF/02 Added util/ratelimit.pl
5330 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
5331 clock is set back in time.
5333 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
5334 Candler <B.Candler@pobox.com>.
5336 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
5337 OK for addresses that are the result of redirections.
5339 PH/48 A number of further additions for the benefit of the new test suite,
5340 including a fake gethostbyname() that interfaces to the fake DNS resolver
5343 TF/06 The fix for widen_domains has also been applied to qualify_single and
5344 search_parents which are the other dnslookup options that can cause
5347 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
5350 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
5352 TF/07 Exim produced the error message "an SRV record indicated no SMTP
5353 service" if it encountered an MX record with an empty target hostname.
5354 The message is now "an MX or SRV record indicated no SMTP service".
5356 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
5357 if the DNS of the sending site is misconfigured. This is quite a
5358 common situation. This change restores the behaviour of treating a
5359 helo verification defer as a failure.
5361 PH/51 If self=fail was set on a router, the bounce message did not include the
5362 actual error message.
5368 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
5370 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
5371 command itself ended in error, the underlying message about the transport
5372 filter timeout was being overwritten with the pipe command error. Now the
5373 underlying error message should be appended to the second error message.
5375 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
5377 PH/02 Reduce the amount of output that "make" produces by default. Full output
5378 can still be requested.
5380 PH/03 The warning log line about a condition test deferring for a "warn" verb
5381 was being output only once per connection, rather than after each
5382 occurrence (because it was using the same function as for successful
5383 "warn" verbs). This seems wrong, so I have changed it.
5385 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
5386 it should not have, which might have caused a crash in the right
5387 circumstances, but probably never did.
5389 PH/04 Installed a modified version of Tony Finch's patch to make submission
5390 mode fix the return path as well as the Sender: header line, and to
5391 add a /name= option so that you can make the user's friendly name appear
5394 TF/03 Added the control = fakedefer ACL modifier.
5396 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
5397 Mark Lowes for thorough testing.
5399 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
5401 TK/03 Merged latest SRS patch from Miles Wilton.
5403 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
5404 with the definition in sysexits.h (which is #included earlier).
5405 Fortunately, Exim does not actually use EX_OK. The code used to try to
5406 preserve the sysexits.h value, by assuming that macro definitions were
5407 scanned for macro replacements. I have been disabused of this notion,
5408 so now the code just undefines EX_OK before #including unistd.h.
5410 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
5411 in the smtp transport. When a block could not be written in a single
5412 write() function, the timeout was being re-applied to each part-write.
5413 This seems wrong - if the receiver was accepting one byte at a time it
5414 would take for ever. The timeout is now adjusted when this happens. It
5415 doesn't have to be particularly precise.
5417 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
5418 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
5420 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
5421 without a domain if the check to <postmaster@domain> fails.
5423 SC/01 Eximstats: added -xls and the ability to specify output files
5424 (patch written by Frank Heydlauf).
5426 SC/02 Eximstats: use FileHandles for outputting results.
5428 SC/03 Eximstats: allow any combination of xls, txt, and html output.
5430 SC/04 Eximstats: fixed display of large numbers with -nvr option
5432 SC/05 Eximstats: fixed merging of reports with empty tables.
5434 SC/06 Eximstats: added the -include_original_destination flag
5436 SC/07 Eximstats: removed tabs and trailing whitespace.
5438 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
5440 TK/06 MBOX spool code: Add real "From " MBOX separator line
5441 so the .eml file is really in mbox format (even though
5442 most programs do not really care). Patch from Alex Miller.
5444 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
5445 The latter is generated from $received_to and is only set if the
5446 message has one envelope recipient. SA can use these headers,
5447 obviously out-of-the-box. Patch from Alex Miller.
5449 PH/08 The ${def test on a variable was returning false if the variable's
5450 value was "0", contrary to what the specification has always said!
5451 The result should be true unless the variable is empty.
5453 PH/09 The syntax error of a character other than { following "${if
5454 def:variable_name" (after optional whitespace) was not being diagnosed.
5455 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
5456 accidental colon was present, for example, could give incorrect results.
5458 PH/10 Tidied the code in a number of places where the st_size field of a stat()
5459 result is used (not including appendfile, where other changes are about
5462 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
5463 This involved changing a lot of size variables from int to off_t. It
5464 should work with maildirs and everything.
5466 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
5467 spamd dying while we are connected to it.
5469 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
5472 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
5473 function for BDB 4.3.
5475 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
5477 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
5478 the sources for PCRE have been re-arranged and more files are now
5481 PH/15 The code I had for printing potentially long long variables in PH/11
5482 above was not the best (it lost precision). The length of off_t variables
5483 is now inspected at build time, and an appropriate printing format (%ld
5484 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
5485 to be "long long int" or "long int". This is needed for the internal
5486 formatting function string_vformat().
5488 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
5489 the configuration file to be ":syslog", then the script "guesses" where
5490 the logs files are, rather than using the compiled in default. In our
5491 case the guess is not the same as the compiled default, so the script
5492 suddenly stopped working when I started to use syslog. The patch checks
5493 to see if log_file_path is "". If so, it attempts to read it from exim
5494 with no configuration file to get the compiled in version, before it
5495 falls back to the previous guessing code."
5497 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
5498 implementing BATV in an Exim configuration. See NewStuff for the gory
5501 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
5502 Makefile that are specific to HP-UX.
5504 PH/18 If the "use_postmaster" option was set for a recipient callout together
5505 with the "random" option, the postmaster address was used as the MAIL
5506 FROM address for the random test, but not for the subsequent recipient
5507 test. It is now used for both.
5509 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
5510 patch removes a few documentation additions to RFC 3028, because the
5511 latest draft now contains them. It adds the new en;ascii-case comparator
5512 and a new error check for 8bit text in MIME parts. Comparator and
5513 require names are now matched exactly. I enabled the subaddress
5514 extension, but it is not well tested yet (read: it works for me)."
5516 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
5517 rework some of the code of TK/09 above to avoid the hardwired use of
5518 "%lld" and "long long". Replaced the call to snprintf() with a call to
5521 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
5522 records point to non-existent hosts", "retry timeout exceeded", and
5523 "retry time not reached for any host after a long failure period".
5525 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
5526 experimental DomainKeys support:
5528 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
5529 (2) On an error such as an illegally used "control", the wrong name for
5530 the control was given.
5532 These problems did NOT occur unless DomainKeys support was compiled.
5534 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
5536 PH/24 Added ${if match_ip condition.
5538 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
5539 databases so that it will be absolutely obvious if a crash occurs in the
5540 DB library. This is a regular occurrence (often caused by mis-matched
5543 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
5544 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
5545 picked up on a user's system that detects such things. There doesn't seem
5546 to be a gcc warning option for this - only an attribute that has to be
5547 put on the function's prototype. It seems that in Fedora Core 4 they have
5548 set this on a number of new functions. No doubt there will be more in due
5551 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
5552 specify a transport. However, if an address that was verified by such a
5553 router was the subject of a callout, Exim crashed because it tried to
5554 read the rcpt_include_affixes from the non-existent transport. Now it
5555 just assumes that the setting of that option is false. This bug was
5556 introduced by 4.51/PH/31.
5558 PH/28 Changed -d+all to exclude +memory, because that information is very
5559 rarely of interest, but it makes the output a lot bigger. People tend to
5560 do -d+all out of habit.
5562 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
5563 code in os-type was giving problems when libc.so lives in lib64, like on
5566 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
5567 aren't the modern standard, and it seems that some systems' include files
5568 don't always have them. Exim was already checking for some of the newer
5569 ones like T_AAAA, and defining it itself. I've added checks for all the
5570 record types that Exim uses.
5572 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
5573 not automatically generating a new one, as it is supposed to. This
5574 prevented TLS from working. If the file did exist, but contained invalid
5575 data, a new version was generated, as expected. It was only the case of a
5576 non-existent file that was broken.
5578 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
5579 with a change in libdomainkeys > 0.64.
5581 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
5582 from DNS. If the selector record carries the flag, it now has
5583 precedence over the domain-wide flag.
5585 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
5587 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
5588 the use of a port name that isn't defined in /etc/services) occurred, the
5589 message was deferred as in a normal delivery, and thus remained on the
5590 spool, instead of being failed because of the mua_wrapper setting. This
5591 is now fixed, and I tidied up some of the mua_wrapper messages at the
5594 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
5595 the messages in a hash of arrays rather than using individual hashes.
5596 This is a bit cleaner and results in dramatic memory savings, albeit
5597 at a slight CPU cost.
5599 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
5600 as requested by Marc Sherman.
5602 SC/10 Eximstats: added histograms for user specified patterns as requested
5605 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
5607 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
5608 fopen() in the content-scanning modules that did not already have it.
5614 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
5615 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
5617 TK/02 Fix ACL "control" statement not being available in MIME ACL.
5619 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
5621 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
5622 to test Sieve filters that use "vacation".
5624 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
5625 that changes the way the GnuTLS parameters are stored in the cache file.
5626 The new format can be generated externally. For backward compatibility,
5627 if the data in the cache doesn't make sense, Exim assumes it has read an
5628 old-format file, and it generates new data and writes a new file. This
5629 means that you can't go back to an older release without removing the
5632 PH/03 A redirect router that has both "unseen" and "one_time" set does not
5633 work if there are any delivery delays because "one_time" forces the
5634 parent to be marked "delivered", so its unseen clone is never tried
5635 again. For this reason, Exim now forbids the simultaneous setting of
5638 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
5639 redirected to themselves ("homonym" addresses). Read the long ChangeLog
5640 entry if you want to know the details. The fix, however, neglected to
5641 consider the case when local delivery batching is involved. The test for
5642 "previously delivered" was not happening when checking to see if an
5643 address could be batched with a previous (undelivered) one; under
5644 certain circumstances this could lead to multiple deliveries to the same
5647 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
5648 in its include files, and this causes problems building Exim.
5650 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
5651 header_syntax) but Exim was just ignoring anything given after a slash.
5652 In particular, this caused confusion with an attempt to use "verify =
5653 reverse_host_lookup/defer_ok". An error is now given when options are
5654 supplied for verify items that do not have them. (Maybe reverse_host_
5655 lookup should have a defer_ok option, but that's a different point.)
5657 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
5658 defined by RFC 821) to 2048, because there were problems with some AUTH
5659 commands, and RFC 1869 says the size should be increased for extended
5660 SMTP commands that take arguments.
5662 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
5665 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
5666 "unknown" error; now it says that the functionality isn't in the binary.
5668 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
5669 an address' error message when a string expansion fails (syntax or
5670 whatever). Otherwise the password may appear in the log. Following change
5671 PH/42 below, there is no longer a chance of it appearing in a bounce
5674 PH/11 Installed exipick version 20050225.0 from John Jetmore.
5676 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
5677 of its addresses was ever tried. (Bugzilla bug #2.)
5679 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
5680 the result incorrectly in the debug output. (It correctly added a newline
5681 to what was transported.)
5683 TF/01 Added $received_time.
5685 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
5686 commented out examples of how to interface to a virus scanner and to
5687 SpamAssassin. Also added commented examples of av_scanner and
5688 spamd_address settings.
5690 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
5691 and controls are allowed in which ACLs. There were a couple of minor
5692 errors. Some of the entries in the conditions table (which is a table of
5693 where they are NOT allowed) were getting very unwieldy; rewrote them as a
5694 negation of where the condition IS allowed.
5696 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
5698 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
5699 header file does not have a version number, so I've had to invent a new
5700 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
5701 API. The code is untested by me (my Linux distribution still has 0.3.2 of
5702 radiusclient), but it was contributed by a Radius user.
5704 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
5705 files or directories, for OpenSSL.
5707 PH/19 When an Exim process that is running as root has to create an Exim log
5708 file, it does so in a subprocess that runs as exim:exim so as to get the
5709 ownership right at creation (otherwise, other Exim processes might see
5710 the file with the wrong ownership). There was no test for failure of this
5711 fork() call, which would lead to the process getting stuck as it waited
5712 for a non-existent subprocess. Forks do occasionally fail when resources
5713 run out. I reviewed all the other calls to fork(); they all seem to check
5716 PH/20 When checking for unexpected SMTP input at connect time (before writing
5717 the banner), Exim was not dealing correctly with a non-positive return
5718 from the read() function. If the client had disconnected by this time,
5719 the result was a log entry for a synchronization error with an empty
5720 string after "input=" when read() returned zero. If read() returned -1
5721 (an event I could not check), uninitialized data bytes were printed.
5722 There were reports of junk text (parts of files, etc) appearing after
5725 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
5727 PH/22 Added support for macro redefinition, and (re)definition in between
5728 driver and ACL definitions.
5730 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
5731 forgetting to use the resulting value; it was using the unexpanded value.
5733 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
5734 hadn't been configured. The fix is from Juergen Kreileder, who
5735 understands it better than I do:
5737 "Here's what I see happening with three configured cyrus_sasl
5738 authenticators configured (plain, login, cram-md5):
5740 On startup auth_cyrus_sasl_init() gets called for each of these.
5741 This means three calls to sasl_listmech() without a specified mech_list.
5742 => SASL tests which mechs of all available mechs actually work
5743 => three warnings about OTP not working
5744 => the returned list contains: plain, login, cram-md5, digest-md5, ...
5746 With the patch, sasl_listmech() also gets called three times. But now
5747 SASL's mech_list option is set to the server_mech specified in the the
5748 authenticator. Or in other words, the answer from sasl_listmech()
5749 gets limited to just the mech you're testing for (which is different
5751 => the return list contains just 'plain' or 'login', 'cram-md5' or
5752 nothing depending on the value of ob->server_mech.
5754 I've just tested the patch: Authentication still works fine,
5755 unavailable mechs specified in the exim configuration are still
5756 caught, and the auth.log warnings about OTP are gone."
5758 PH/25 When debugging is enabled, the contents of the command line are added
5759 to the debugging output, even when log_selector=+arguments is not
5762 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
5763 answer is "GNU", and only if the return is "GNU/something" is the answer
5766 PH/27 $acl_verify_message is now set immediately after the failure of a
5767 verification in an ACL, and so is available in subsequent modifiers. In
5768 particular, the message can be preserved by coding like this:
5770 warn !verify = sender
5771 set acl_m0 = $acl_verify_message
5773 Previously, $acl_verify_message was set only while expanding "message"
5774 and "log_message" when a very denied access.
5776 PH/28 Modified OS/os.c-Linux with
5778 -#ifndef OS_LOAD_AVERAGE
5779 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
5781 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
5782 nomenclature these days.)
5784 PH/29 Installed patch from the Sieve maintainer that adds the options
5785 sieve_useraddress and sieve_subaddress to the redirect router.
5787 PH/30 In these circumstances:
5788 . Two addresses routed to the same list of hosts;
5789 . First host does not offer TLS;
5790 . First host accepts first address;
5791 . First host gives temporary error to second address;
5792 . Second host offers TLS and a TLS session is established;
5793 . Second host accepts second address.
5794 Exim incorrectly logged both deliveries with the TLS parameters (cipher
5795 and peerdn, if requested) that were in fact used only for the second
5798 PH/31 When doing a callout as part of verifying an address, Exim was not paying
5799 attention to any local part prefix or suffix that was matched by the
5800 router that accepted the address. It now behaves in the same way as it
5801 does for delivery: the affixes are removed from the local part unless
5802 rcpt_include_affixes is set on the transport.
5804 PH/32 Add the sender address, as F=<...>, to the log line when logging a
5805 timeout during the DATA phase of an incoming message.
5807 PH/33 Sieve envelope tests were broken for match types other than :is. I have
5808 applied a patch sanctioned by the Sieve maintainer.
5810 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
5811 the uid or gid is negative. A case of a negative gid caused this to be
5812 noticed. The fix allows for either to be negative.
5814 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
5815 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
5818 PH/36 The change PH/12 above was broken. Fixed it.
5820 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
5821 the grounds that routing the same address twice would always produce the
5822 same answer. This might have been true once, but it is certainly no
5823 longer true now. Routing a child address may depend on the previous
5824 routing that produced that child. Some complicated redirection strategies
5825 went wrong when messages had multiple recipients, and made Exim's
5826 behaviour dependent on the order in which the addresses were given.
5828 I have moved the duplicate checking until after the routing is complete.
5829 Exim scans the addresses that are assigned to local and remote
5830 transports, and removes any duplicates. This means that more work will be
5831 done, as duplicates will always all be routed, but duplicates are
5832 presumably rare, so I don't expect this is of any significance.
5834 For deliveries to pipes, files, and autoreplies, the duplicate checking
5835 still happens during the routing process, since they are not going to be
5838 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
5839 It corrects a timeout issue with spamd. This is Ian's comment: "The
5840 background is that sometimes spamd either never reads data from a
5841 connection it has accepted, or it never writes response data. The exiscan
5842 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
5843 blindly assumes that writes won't block so it may never time out."
5845 PH/39 Allow G after quota size as well as K and M.
5847 PH/40 The value set for $authenticated_id in an authenticator may not contain
5848 binary zeroes or newlines because the value is written to log lines and
5849 to spool files. There was no check on this. Now the value is run through
5850 the string_printing() function so that such characters are converted to
5851 printable escape sequences.
5853 PH/41 $message_linecount is a new variable that contains the total number of
5854 lines in the message. Compare $body_linecount, which is the count for the
5857 PH/42 Exim no longer gives details of delivery errors for specific addresses in
5858 bounce and delay warning messages, except in certain special cases, which
5861 (a) An SMTP error message from a remote host;
5862 (b) A message specified in a :fail: redirection;
5863 (c) A message specified in a "fail" command in a system filter;
5864 (d) A message specified in a FAIL return from the queryprogram router;
5865 (e) A message specified by the cannot_route_message router option.
5867 In these cases only, Exim does include the error details in bounce and
5868 warning messages. There are also a few cases where bland messages such
5869 as "unrouteable address" or "local delivery error" are given.
5871 PH/43 $value is now also set for the "else" part of a ${run expansion.
5873 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
5874 being worked on, but at least Exim now implements the latest version to
5877 PH/45 In a pipe transport, although a timeout while waiting for the pipe
5878 process to complete was treated as a delivery failure, a timeout while
5879 writing the message to the pipe was logged, but erroneously treated as a
5880 successful delivery. Such timeouts include transport filter timeouts. For
5881 consistency with the overall process timeout, these timeouts are now
5882 treated as errors, giving rise to delivery failures by default. However,
5883 there is now a new Boolean option for the pipe transport called
5884 timeout_defer, which, if set TRUE, converts the failures into defers for
5885 both kinds of timeout. A transport filter timeout is now identified in
5888 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
5889 systems where "make" and "gmake" are different, calling "gmake" at top
5890 level broke things. I've arranged for the value of $(MAKE) to be passed
5891 from the Makefile to this script so that it can call the same version of
5895 A note about Exim versions 4.44 and 4.50
5896 ----------------------------------------
5898 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
5899 changes of various kinds. As a consequence, a big documentation update was
5900 needed. This delayed the release for rather longer than seemed good, especially
5901 in the light of a couple of (minor) security issues. Therefore, the changes
5902 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
5903 release. So 4.44 and 4.50 are in effect two different branches that both start
5906 I have left the 4.50 change log unchanged; it contains all the changes since
5907 4.43. The change log for 4.44 is below; many of its items are identical to
5908 those for 4.50. This seems to be the most sensible way to preserve the
5909 historical information.
5915 1. Minor wording change to the doc/README.SIEVE file.
5917 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
5918 computation of the current number of files was incorrect.
5920 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
5921 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
5924 4. Give more explanation in the error message when the command for a transport
5925 filter fails to execute.
5927 5. There are several places where Exim runs a non-Exim command in a
5928 subprocess. The SIGUSR1 signal should be disabled for these processes. This
5929 was being done only for the command run by the queryprogram router. It is
5930 now done for all such subprocesses. The other cases are: ${run, transport
5931 filters, and the commands run by the lmtp and pipe transports.
5933 6. Added CONFIGURE_GROUP build-time option.
5935 7. Some older OS have a limit of 256 on the maximum number of file
5936 descriptors. Exim was using setrlimit() to set 1000 as a large value
5937 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
5938 systems. I've change it so that if it can't get 1000, it tries for 256.
5940 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
5941 was an oversight, and furthermore, ever since the addition of extra
5942 controls (e.g. 4.43/32), the checks on when to allow different forms of
5943 "control" were broken. There should now be diagnostics for all cases when a
5944 control that does not make sense is encountered.
5946 9. Added the /retain_sender option to "control=submission".
5948 10. $recipients is now available in the predata ACL (oversight).
5950 11. Tidy the search cache before the fork to do a delivery from a message
5951 received from the command line. Otherwise the child will trigger a lookup
5952 failure and thereby defer the delivery if it tries to use (for example) a
5953 cached ldap connection that the parent has called unbind on.
5955 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
5956 of $address_data from the recipient verification was clobbered by the
5957 sender verification.
5959 13. The value of address_data from a sender verification is now available in
5960 $sender_address_data in subsequent conditions in the ACL statement.
5962 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
5964 15. Added a new option "connect=<time>" to callout options, to set a different
5967 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
5968 was its contents. (It was OK if the option was not defined at all.)
5970 17. A "Completed" log line is now written for messages that are removed from
5971 the spool by the -Mrm option.
5973 18. New variables $sender_verify_failure and $recipient_verify_failure contain
5974 information about exactly what failed.
5976 19. Added -dd to debug only the daemon process.
5978 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
5979 handles timeouts, both on the server side and network timeouts. Renamed the
5980 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
5982 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
5983 It is now set to "smtps".
5985 22. $host_address is now set to the target address during the checking of
5986 ignore_target_hosts.
5988 23. When checking ignore_target_hosts for an ipliteral router, no host name was
5989 being passed; this would have caused $sender_host_name to have been used if
5990 matching the list had actually called for a host name (not very likely,
5991 since this list is usually IP addresses). A host name is now passed as
5994 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
5995 code that specifies a non-restarting handler (typically sigaction() in
5996 modern systems) in an attempt to fix a rare and obscure crash bug.
5998 25. Narrowed the window for a race in the daemon that could cause it to ignore
5999 SIGCHLD signals. This is not a major problem, because they are used only to
6000 wake it up if nothing else does.
6002 26. A malformed maildirsize file could cause Exim to calculate negative values
6003 for the mailbox size or file count. Odd effects could occur as a result.
6004 The maildirsize information is now recalculated if the size or filecount
6007 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6008 support for a long time. Removed HAVE_SYS_VFS_H.
6010 28. Installed the latest version of exipick from John Jetmore.
6012 29. In an address list, if the pattern was not a regular expression, an empty
6013 subject address (from a bounce message) matched only if the pattern was an
6014 empty string. Non-empty patterns were not even tested. This was the wrong
6015 because it is perfectly reasonable to use an empty address as part of a
6016 database query. An empty address is now tested by patterns that are
6017 lookups. However, all the other forms of pattern expect the subject to
6018 contain a local part and a domain, and therefore, for them, an empty
6019 address still always fails if the pattern is not itself empty.
6021 30. Exim went into a mad DNS loop when attempting to do a callout where the
6022 host was specified on an smtp transport, and looking it up yielded more
6023 than one IP address.
6025 31. Re-factored the code for checking spool and log partition space into a
6026 function that finds that data and another that does the check. The former
6027 is then used to implement four new variables: $spool_space, $log_space,
6028 $spool_inodes, and $log_inodes.
6030 32. The RFC2047 encoding function was originally intended for short strings
6031 such as real names; it was not keeping to the 75-character limit for
6032 encoded words that the RFC imposes. It now respects the limit, and
6033 generates multiple encoded words if necessary. To be on the safe side, I
6034 have increased the buffer size for the ${rfc2047: expansion operator from
6037 33. It is now permitted to omit both strings after an "if" condition; if the
6038 condition is true, the result is "true". As before, when the second string
6039 is omitted, a false condition yields an empty string. This makes it less
6040 cumbersome to write custom ACL and router conditions.
6042 34. Failure to deliver a bounce message always caused it to be frozen, even if
6043 there was an errors_to setting on the router. The errors_to setting is now
6046 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6047 canonical form (fully expanded) before being placed in
6048 $sender_host_address.
6050 36. The table in the code that translates DNS record types into text (T_A to
6051 "A" for instance) was missing entries for NS and CNAME. It is just possible
6052 that this could have caused confusion if both these types were looked up
6053 for the same domain, because the text type is used as part of Exim's
6054 per-process caching. But the chance of anyone hitting this buglet seems
6057 37. The dnsdb lookup has been extended in a number of ways.
6059 (1) There is a new type, "zns", which walks up the domain tree until it
6060 finds some nameserver records. It should be used with care.
6062 (2) There is a new type, "mxh", which is like "mx" except that it returns
6063 just the host names, not the priorities.
6065 (3) It is now possible to give a list of domains (or IP addresses) to be
6066 looked up. The behaviour when one of the lookups defers can be
6067 controlled by a keyword.
6069 (4) It is now possible to specify the separator character for use when
6070 multiple records are returned.
6072 38. The dnslists ACL condition has been extended: it is now possible to supply
6073 a list of IP addresses and/or domains to be looked up in a particular DNS
6076 39. Added log_selector=+queue_time_overall.
6078 40. When running the queue in the test harness, wait just a tad after forking a
6079 delivery process, to get repeatability of debugging output.
6081 41. Include certificate and key file names in error message when GnuTLS fails
6082 to set them up, because the GnuTLS error message doesn't include the name
6083 of the failing file when there is a problem reading it.
6085 42. Allow both -bf and -bF in the same test run.
6087 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
6089 44. The "Exiscan patch" is now merged into the mainline Exim source.
6091 45. Sometimes the final signoff response after QUIT could fail to get
6092 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6093 < 0 before doing a fflush(). This bug looks as though it goes back to the
6094 introduction of TLS in release 3.20, but "sometimes" must have been rare
6095 because the tests only now provoked it.
6097 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
6098 (this can affect the format of dates).
6100 47. exim_tidydb, when checking for the continued existence of a message for
6101 which it has found a message-specific retry record, was not finding
6102 messages that were in split spool directories. Consequently, it was
6103 deleting retry records that should have stayed in existence.
6105 48. Steve fixed some bugs in eximstats.
6107 49. The SPA authentication driver was not abandoning authentication and moving
6108 on to the next authenticator when an expansion was forced to fail,
6109 contradicting the general specification for all authenticators. Instead it
6110 was generating a temporary error. It now behaves as specified.
6112 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
6113 (the order specifies the preference for clients). The order is now AES256,
6114 AES128, 3DES, ARCFOUR128.
6116 51. Small patch to Sieve code - explicitly set From: when generating an
6119 52. Exim crashed if a remote delivery caused a very long error message to be
6120 recorded - for instance if somebody sent an entire SpamAssassin report back
6121 as a large number of 550 error lines. This bug was coincidentally fixed by
6122 increasing the size of one of Exim's internal buffers (big_buffer) that
6123 happened as part of the Exiscan merge. However, to be on the safe side, I
6124 have made the code more robust (and fixed the comments that describe what
6127 53. Now that there can be additional text after "Completed" in log lines (if
6128 the queue_time_overall log selector is set), a one-byte patch to exigrep
6129 was needed to allow it to recognize "Completed" as not the last thing in
6132 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
6133 patch that reportedly fixes this has been added. I am not expert enough to
6134 create a test for it. This is what the patch creator wrote:
6136 "I found a little strange behaviour of ldap code when working with
6137 Windows 2003 AD Domain, where users was placed in more than one
6138 Organization Units. When I tried to give exim partial DN, the exit code
6139 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
6140 But simultaneously result of request was absolutely normal ldap result,
6141 so I produce this patch..."
6143 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
6144 REFERENCE, so I have modified the code to exclude the patch when that macro
6147 55. Some experimental protocols are using DNS PTR records for new purposes. The
6148 keys for these records are domain names, not reversed IP addresses. The
6149 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6150 leaves it alone. Component reversal etc. now happens only for IP addresses.
6153 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
6155 57. Double the size of the debug message buffer (to 2048) so that more of very
6156 long debug lines gets shown.
6158 58. The exicyclog utility now does better if the number of log files to keep
6159 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
6161 59. Two changes related to the smtp_active_hostname option:
6163 (1) $smtp_active_hostname is now available as a variable.
6164 (2) The default for smtp_banner uses $smtp_active_hostname instead
6165 of $primary_hostname.
6167 60. The host_aton() function is supposed to be passed a string that is known
6168 to be a valid IP address. However, in the case of IPv6 addresses, it was
6169 not checking this. This is a hostage to fortune. Exim now panics and dies
6170 if the condition is not met. A case was found where this could be provoked
6171 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6172 components; fortuitously, this particular loophole had already been fixed
6173 by change 4.50/55 above.
6175 If there are any other similar loopholes, the new check in host_aton()
6176 itself should stop them being exploited. The report I received stated that
6177 data on the command line could provoke the exploit when Exim was running as
6178 exim, but did not say which command line option was involved. All I could
6179 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6180 running as the user.
6183 61. There was a buffer overflow vulnerability in the SPA authentication code
6184 (which came originally from the Samba project). I have added a test to the
6185 spa_base64_to_bits() function which I hope fixes it.
6188 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
6189 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
6191 63. The daemon start-up calls getloadavg() while still root for those OS that
6192 need the first call to be done as root, but it missed one case: when
6193 deliver_queue_load_max is set with deliver_drop_privilege. This is
6194 necessary for the benefit of the queue runner, because there is no re-exec
6195 when deliver_drop_privilege is set.
6197 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
6198 This has been fixed.
6200 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6201 was in use, was not putting the data itself into the right store pool;
6202 consequently, it could be overwritten for a subsequent message in the same
6203 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6206 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
6208 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
6209 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
6210 the function were treating the return as a boolean value, which happened to
6211 work because 0=false and not-0=true, but is not correct code.
6213 68. The host_aton() function was not handling scoped IPv6 addresses (those
6214 with, for example, "%eth0" on the end) correctly.
6216 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
6217 negated items (that is, ~something) in unsigned ints. Some compilers
6218 apparently mutter when there is no cast.
6220 70. If an address verification called from an ACL failed, and did not produce a
6221 user-specific message (i.e. there was only a "system" message), nothing was
6222 put in $acl_verify_message. In this situation, it now puts the system
6225 71. Change 4.23/11 added synchronization checking at the start of an SMTP
6226 session; change 4.31/43 added the unwanted input to the log line - except
6227 that it did not do this in the start of session case. It now does.
6229 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
6230 This is wrong and can cause the other end to generate a synchronization
6231 error if it is another Exim or anything else that does the synchronization
6232 check. A QUIT command is no longer sent after a timeout.
6234 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
6235 during host lookups.
6237 74. The defer_ok option of callout verification was not working if it was used
6238 when verifying addresses in header lines, that is, for this case:
6240 verify = header_sender/callout=defer_ok
6242 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
6243 those file descriptors could be used for SMTP connections. If anything
6244 wrote to stderr (the example that came up was "warn" in embedded Perl), it
6245 could be sent to the SMTP client, causing chaos. The daemon now opens
6246 stdin, stdout, and stderr to /dev/null when it puts itself into the
6249 76. Arrange for output from Perl's "warn" command to be written to Exim's main
6250 log by default. The user can override this with suitable Perl magic.
6252 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
6253 the log message when discard triggers, was not working for the DATA ACL or
6254 for the non-SMTP ACL.
6256 78. Error message wording change in sieve.c.
6258 79. If smtp_accept_max_per_host was set, the number of connections could be
6259 restricted to fewer than expected, because the daemon was trying to set up
6260 a new connection before checking whether the processes handling previous
6261 connections had finished. The check for completed processes is now done
6262 earlier. On busy systems, this bug wouldn't be noticed because something
6263 else would have woken the daemon, and it would have reaped the completed
6266 80. If a message was submitted locally by a user whose login name contained one
6267 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
6268 It caused a spool format error. I have fixed the spool reading code. A
6269 related problem was that the "from" clause in the Received: line became
6270 illegal because of the space(s). It is now covered by ${quote_local_part.
6272 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
6275 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
6277 83. Patch from Sieve maintainer to fix unterminated string problem in
6278 "vacation" handling.
6280 84. Some minor changes to the Linux configuration files to help with other
6281 OS variants using glibc.
6283 85. One more patch for Sieve to update vacation handling to latest spec.
6286 ----------------------------------------------------
6287 See the note above about the 4.44 and 4.50 releases.
6288 ----------------------------------------------------
6294 1. Change 4.43/35 introduced a bug that caused file counts to be
6295 incorrectly computed when quota_filecount was set in an appendfile
6298 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
6299 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
6302 3. Give more explanation in the error message when the command for a transport
6303 filter fails to execute.
6305 4. There are several places where Exim runs a non-Exim command in a
6306 subprocess. The SIGUSR1 signal should be disabled for these processes. This
6307 was being done only for the command run by the queryprogram router. It is
6308 now done for all such subprocesses. The other cases are: ${run, transport
6309 filters, and the commands run by the lmtp and pipe transports.
6311 5. Some older OS have a limit of 256 on the maximum number of file
6312 descriptors. Exim was using setrlimit() to set 1000 as a large value
6313 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
6314 systems. I've change it so that if it can't get 1000, it tries for 256.
6316 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
6317 was an oversight, and furthermore, ever since the addition of extra
6318 controls (e.g. 4.43/32), the checks on when to allow different forms of
6319 "control" were broken. There should now be diagnostics for all cases when a
6320 control that does not make sense is encountered.
6322 7. $recipients is now available in the predata ACL (oversight).
6324 8. Tidy the search cache before the fork to do a delivery from a message
6325 received from the command line. Otherwise the child will trigger a lookup
6326 failure and thereby defer the delivery if it tries to use (for example) a
6327 cached ldap connection that the parent has called unbind on.
6329 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
6330 of $address_data from the recipient verification was clobbered by the
6331 sender verification.
6333 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
6334 was its contents. (It was OK if the option was not defined at all.)
6336 11. A "Completed" log line is now written for messages that are removed from
6337 the spool by the -Mrm option.
6339 12. $host_address is now set to the target address during the checking of
6340 ignore_target_hosts.
6342 13. When checking ignore_target_hosts for an ipliteral router, no host name was
6343 being passed; this would have caused $sender_host_name to have been used if
6344 matching the list had actually called for a host name (not very likely,
6345 since this list is usually IP addresses). A host name is now passed as
6348 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
6349 code that specifies a non-restarting handler (typically sigaction() in
6350 modern systems) in an attempt to fix a rare and obscure crash bug.
6352 15. Narrowed the window for a race in the daemon that could cause it to ignore
6353 SIGCHLD signals. This is not a major problem, because they are used only to
6354 wake it up if nothing else does.
6356 16. A malformed maildirsize file could cause Exim to calculate negative values
6357 for the mailbox size or file count. Odd effects could occur as a result.
6358 The maildirsize information is now recalculated if the size or filecount
6361 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
6362 support for a long time. Removed HAVE_SYS_VFS_H.
6364 18. Updated exipick to current release from John Jetmore.
6366 19. Allow an empty sender to be matched against a lookup in an address list.
6367 Previously the only cases considered were a regular expression, or an
6370 20. Exim went into a mad DNS lookup loop when doing a callout where the
6371 host was specified on the transport, if the DNS lookup yielded more than
6374 21. The RFC2047 encoding function was originally intended for short strings
6375 such as real names; it was not keeping to the 75-character limit for
6376 encoded words that the RFC imposes. It now respects the limit, and
6377 generates multiple encoded words if necessary. To be on the safe side, I
6378 have increased the buffer size for the ${rfc2047: expansion operator from
6381 22. Failure to deliver a bounce message always caused it to be frozen, even if
6382 there was an errors_to setting on the router. The errors_to setting is now
6385 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
6386 canonical form (fully expanded) before being placed in
6387 $sender_host_address.
6389 24. Updated eximstats to version 1.33
6391 25. Include certificate and key file names in error message when GnuTLS fails
6392 to set them up, because the GnuTLS error message doesn't include the name
6393 of the failing file when there is a problem reading it.
6395 26. Expand error message when OpenSSL has problems setting up cert/key files.
6398 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
6399 (this can affect the format of dates).
6401 28. exim_tidydb, when checking for the continued existence of a message for
6402 which it has found a message-specific retry record, was not finding
6403 messages that were in split spool directories. Consequently, it was
6404 deleting retry records that should have stayed in existence.
6406 29. eximstats updated to version 1.35
6407 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
6408 1.35 - bugfix such that pie charts by volume are generated correctly
6410 30. The SPA authentication driver was not abandoning authentication and moving
6411 on to the next authenticator when an expansion was forced to fail,
6412 contradicting the general specification for all authenticators. Instead it
6413 was generating a temporary error. It now behaves as specified.
6415 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
6416 (the order specifies the preference for clients). The order is now AES256,
6417 AES128, 3DES, ARCFOUR128.
6419 31. Small patch to Sieve code - explicitly set From: when generating an
6422 32. Exim crashed if a remote delivery caused a very long error message to be
6423 recorded - for instance if somebody sent an entire SpamAssassin report back
6424 as a large number of 550 error lines. This bug was coincidentally fixed by
6425 increasing the size of one of Exim's internal buffers (big_buffer) that
6426 happened as part of the Exiscan merge. However, to be on the safe side, I
6427 have made the code more robust (and fixed the comments that describe what
6430 33. Some experimental protocols are using DNS PTR records for new purposes. The
6431 keys for these records are domain names, not reversed IP addresses. The
6432 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
6433 leaves it alone. Component reversal etc. now happens only for IP addresses.
6436 34. The host_aton() function is supposed to be passed a string that is known
6437 to be a valid IP address. However, in the case of IPv6 addresses, it was
6438 not checking this. This is a hostage to fortune. Exim now panics and dies
6439 if the condition is not met. A case was found where this could be provoked
6440 from a dnsdb PTR lookup with an IPv6 address that had more than 8
6441 components; fortuitously, this particular loophole had already been fixed
6442 by change 4.50/55 or 4.44/33 above.
6444 If there are any other similar loopholes, the new check in host_aton()
6445 itself should stop them being exploited. The report I received stated that
6446 data on the command line could provoke the exploit when Exim was running as
6447 exim, but did not say which command line option was involved. All I could
6448 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
6449 running as the user.
6452 35. There was a buffer overflow vulnerability in the SPA authentication code
6453 (which came originally from the Samba project). I have added a test to the
6454 spa_base64_to_bits() function which I hope fixes it.
6457 36. The daemon start-up calls getloadavg() while still root for those OS that
6458 need the first call to be done as root, but it missed one case: when
6459 deliver_queue_load_max is set with deliver_drop_privilege. This is
6460 necessary for the benefit of the queue runner, because there is no re-exec
6461 when deliver_drop_privilege is set.
6463 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
6464 was in use, was not putting the data itself into the right store pool;
6465 consequently, it could be overwritten for a subsequent message in the same
6466 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
6469 38. Sometimes the final signoff response after QUIT could fail to get
6470 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
6471 < 0 before doing a fflush(). This bug looks as though it goes back to the
6472 introduction of TLS in release 3.20, but "sometimes" must have been rare
6473 because the tests only now provoked it.
6479 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
6480 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
6481 Now it returns an integer. A number of calls were still expecting a T/F
6482 return. Fortuitously, in all cases, the tests worked in OK situations,
6483 which is the norm. However, things would have gone wrong on any write
6484 failures on the smtp file descriptor. This function is used when sending
6485 messages over SMTP and also when doing verify callouts.
6487 2. When Exim is called to do synchronous delivery of a locally submitted
6488 message (the -odf or -odi options), it no longer closes stderr before doing
6491 3. Implemented the mua_wrapper option.
6493 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
6495 5. Implemented the functions header_remove(), header_testname(),
6496 header_add_at_position(), and receive_remove_recipient(), and exported them
6499 6. If an ACL "warn" statement specified the addition of headers, Exim already
6500 inserted X-ACL-Warn: at the start if there was no header name. However, it
6501 was not making this test for the second and subsequent header lines if
6502 there were newlines in the string. This meant that an invalid header could
6503 be inserted if Exim was badly configured.
6505 7. Allow an ACL "warn" statement to add header lines at the start or after all
6506 the Received: headers, as well as at the end.
6508 8. Added the rcpt_4xx retry error code.
6510 9. Added postmaster_mailfrom=xxx to callout verification option.
6512 10. Added mailfrom=xxxx to the callout verification option, for verify=
6515 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
6516 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
6518 12. Inserted some casts to stop certain compilers warning when using pointer
6519 differences as field lengths or precisions in printf-type calls (mostly
6520 affecting debugging statements).
6522 13. Added optional readline() support for -be (dynamically loaded).
6524 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
6525 same clock tick as a message's arrival, so that its received time was the
6526 same as the "first fail" time on the retry record, and that message
6527 remained on the queue past the ultimate address timeout, every queue runner
6528 would try a delivery (because it was past the ultimate address timeout) but
6529 after another failure, the ultimate address timeout, which should have then
6530 bounced the address, did not kick in. This was a "< instead of <=" error;
6531 in most cases the first failure would have been in the next clock tick
6532 after the received time, and all would be well.
6534 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
6535 being recognized when the domain list was tested by the match_domain
6536 condition in an expansion string.
6538 16. Added the ${str2b64: operator.
6540 17. Exim was always calling setrlimit() to set a large limit for the number of
6541 processes, without checking whether the existing limit was already
6542 adequate. (It did check for the limit on file descriptors.) Furthermore,
6543 errors from getrlimit() and setrlimit() were being ignored. Now they are
6544 logged to the main and panic logs, but Exim does carry on, to try to do its
6545 job under whatever limits there are.
6547 18. Imported PCRE 5.0.
6549 19. Trivial typo in log message " temporarily refused connection" (the leading
6552 20. If the log selector return_path_on_delivery was set and an address was
6553 redirected to /dev/null, the delivery process crashed because it assumed
6554 that a return path would always be set for a "successful" delivery. In this
6555 case, the whole delivery is bypassed as an optimization, and therefore no
6558 21. Internal re-arrangement: the function for sending a challenge and reading
6559 a response while authentication was assuming a zero-terminated challenge
6560 string. It's now changed to take a pointer and a length, to allow for
6561 binary data in such strings.
6563 22. Added the cyrus_sasl authenticator (code supplied by MBM).
6565 23. Exim was not respecting finduser_retries when seeking the login of the
6566 uid under which it was called; it was always trying 10 times. (The default
6567 setting of finduser_retries is zero.) Also, it was sleeping after the final
6568 failure, which is pointless.
6570 24. Implemented tls_on_connect_ports.
6572 25. Implemented acl_smtp_predata.
6574 26. If the domain in control=submission is set empty, Exim assumes that the
6575 authenticated id is a complete email address when it generates From: or
6576 Sender: header lines.
6578 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
6579 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
6580 chown and chgrp in /bin and hostname in /usr/bin.
6582 28. Exim was keeping the "process log" file open after each use, just as it
6583 does for the main log. This opens the possibility of it remaining open for
6584 long periods when the USR1 signal hits a daemon. Occasional processlog
6585 errors were reported, that could have been caused by this. Anyway, it seems
6586 much more sensible not to leave this file open at all, so that is what now
6589 29. The long-running daemon process does not normally write to the log once it
6590 has entered its main loop, and it closes the log before doing so. This is
6591 so that log files can straightforwardly be renamed and moved. However,
6592 there are a couple of unusual error situations where the daemon does write
6593 log entries, and I had neglected to close the log afterwards.
6595 30. The text of an SMTP error response that was received during a remote
6596 delivery was being truncated at 512 bytes. This is too short for some of
6597 the long messages that one sometimes sees. I've increased the limit to
6600 31. It is now possible to make retry rules that apply only when a message has a
6601 specific sender, in particular, an empty sender.
6603 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
6604 it possible to be selective about when SMTP synchronization is enforced.
6606 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
6608 32. Implemented hosts_connection_nolog.
6610 33. Added an ACL for QUIT.
6612 34. Setting "delay_warning=" to disable warnings was not working; it gave a
6615 35. Added mailbox_size and mailbox_filecount to appendfile.
6617 36. Added control = no_multiline_responses to ACLs.
6619 37. There was a bug in the logic of the code that waits for the clock to tick
6620 in the case where the clock went backwards by a substantial amount such
6621 that the microsecond fraction of "now" was more than the microsecond
6622 fraction of "then" (but the whole seconds number was less).
6624 38. Added support for the libradius Radius client library this is found on
6625 FreeBSD (previously only the radiusclient library was supported).
6631 1. When certain lookups returned multiple values in the form name=value, the
6632 quoting of the values was not always being done properly. Specifically:
6633 (a) If the value started with a double quote, but contained no whitespace,
6635 (b) If the value contained whitespace other than a space character (i.e.
6636 tabs or newlines or carriage returns) it was not quoted.
6637 This fix has been applied to the mysql and pgsql lookups by writing a
6638 separate quoting function and calling it from the lookup code. The fix
6639 should probably also be applied to nisplus, ibase and oracle lookups, but
6640 since I cannot test any of those, I have not disturbed their existing code.
6642 2. A hit in the callout cache for a specific address caused a log line with no
6643 reason for rejecting RCPT. Now it says "Previous (cached) callout
6644 verification failure".
6646 3. There was an off-by-one bug in the queryprogram router. An over-long
6647 return line was truncated at 256 instead of 255 characters, thereby
6648 overflowing its buffer with the terminating zero. As well as fixing this, I
6649 have increased the buffer size to 1024 (and made a note to document this).
6651 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
6652 when Exim is waiting for an SMTP response from a remote server, Exim
6653 restarts its select() call on the socket, thereby resetting its timeout.
6654 This is not a problem when such interrupts are rare. Somebody set up a cron
6655 job to run exiwhat every 2 minutes, which is less than the normal select()
6656 timeout (5 or 10 minutes). This meant that the select() timeout never
6657 kicked in because it was always reset. I have fixed this by comparing the
6658 time when an interrupt arrives with the time at the start of the first call
6659 to select(). If more time than the timeout has elapsed, the interrupt is
6660 treated as a timeout.
6662 5. Some internal re-factoring in preparation for the addition of Sieve
6663 extensions (by MH). In particular, the "personal" test is moved to a
6664 separate function, and given an option for scanning Cc: and Bcc: (which is
6665 not set for Exim filters).
6667 6. When Exim created an email address using the login of the caller as the
6668 local part (e.g. when creating a From: or Sender: header line), it was not
6669 quoting the local part when it contained special characters such as @.
6671 7. Installed new OpenBSD configuration files.
6673 8. Reworded some messages for syntax errors in "and" and "or" conditions to
6674 try to make them clearer.
6676 9. Callout options, other than the timeout value, were being ignored when
6677 verifying sender addresses in header lines. For example, when using
6679 verify = header_sender/callout=no_cache
6681 the cache was (incorrectly) being used.
6683 10. Added a missing instance of ${EXE} to the exim_install script; this affects
6684 only the Cygwin environment.
6686 11. When return_path_on_delivery was set as a log selector, if different remote
6687 addresses in the same message used different return paths and parallel
6688 remote delivery occurred, the wrong values would sometimes be logged.
6689 (Whenever a remote delivery process finished, the return path value from
6690 the most recently started remote delivery process was logged.)
6692 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
6693 lines when AUTH and/or TLS are in use. This is the "received protocol"
6694 field. Exim used to use "asmtp" for authenticated SMTP, without any
6695 indication (in the protocol name) for TLS use. Now it follows the RFC and
6696 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
6697 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
6698 names appear in log lines as well as in Received: header lines.
6700 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
6701 extensions, and comparison tests, and to fix some bugs.
6703 14. Changes to the "personal" filter test:
6705 (1) The test was buggy in that it was just doing the equivalent of
6706 "contains" tests on header lines. For example, if a user's address was
6707 anne@some.where, the "personal" test would incorrectly be true for
6709 To: susanne@some.where
6711 This test is now done by extracting each address from the header in turn,
6712 and checking the entire address. Other tests that are part of "personal"
6713 are now done using regular expressions (for example, to check local parts
6714 of addresses in From: header lines).
6716 (2) The list of non-personal local parts in From: addresses has been
6717 extended to include "listserv", "majordomo", "*-request", and "owner-*",
6718 taken from the Sieve specification recommendations.
6720 (3) If the message contains any header line starting with "List-" it is
6721 treated as non-personal.
6723 (4) The test for "circular" in the Subject: header line has been removed
6724 because it now seems ill-conceived.
6726 15. Minor typos in src/EDITME comments corrected.
6728 16. Installed latest exipick from John Jetmore.
6730 17. If headers_add on a router specified a text string that was too long for
6731 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
6732 of string_sprintf() is now avoided.
6734 18. $message_body_size was not set (it was always zero) when running the DATA
6735 ACL and the local_scan() function.
6737 19. For the "mail" command in an Exim filter, no default was being set for
6738 the once_repeat time, causing a random time value to be used if "once" was
6739 specified. (If the value happened to be <= 0, no repeat happened.) The
6740 default is now 0s, meaning "never repeat". The "vacation" command was OK
6741 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
6742 (I found it when inspecting the code).
6744 20. There is now an overall timeout for performing a callout verification. It
6745 defaults to 4 times the callout timeout, which applies to individual SMTP
6746 commands during the callout. The overall timeout applies when there is more
6747 than one host that can be tried. The timeout is checked before trying the
6748 next host. This prevents very long delays if there are a large number of
6749 hosts and all are timing out (e.g. when the network connections are timing
6750 out). The value of the overall timeout can be changed by specifying an
6751 additional sub-option for "callout", called "maxwait". For example:
6753 verify = sender/callout=5s,maxwait=20s
6755 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
6756 to the end before writing, but this should make it even safer).
6758 22. Exim was forgetting that it had advertised PIPELINING for the second and
6759 subsequent messages on an SMTP connection. It was also not resetting its
6760 memory on STARTTLS and an internal HELO.
6762 23. When Exim logs an SMTP synchronization error within a session, it now
6763 records whether PIPELINING has been advertised or not.
6765 24. Added 3 instances of "(long int)" casts to time_t variables that were being
6766 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
6767 rather than long int.
6769 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
6771 26. Added the never_mail option to autoreply.
6777 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
6778 crash if the getsockname() call failed; this can happen if a connection is
6779 closed very soon after it is established. The problem was simply in the
6780 order in which certain operations were done, causing Exim to try to write
6781 to the SMTP stream before it had set up the file descriptor. The bug has
6782 been fixed by making things happen in the correct order.
6788 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
6789 before the connection was closed, thus losing the rejection response.
6791 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
6792 some early Solaris releases, but causes trouble in current releases where
6793 socklen_t is defined.
6795 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
6798 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
6801 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
6802 that releasing the top of it at the end releases what was used for sub-
6803 expansions (unless the block got too big). However, discard this block if
6804 the first thing is a variable or header, so that we can use its block when
6805 it is dynamic (useful for very large $message_headers, for example).
6807 6. Lookups now cache *every* query, not just the most recent. A new, separate
6808 store pool is used for this. It can be recovered when all lookup caches are
6809 flushed. Lookups now release memory at the end of their result strings.
6810 This has involved some general refactoring of the lookup sources.
6812 7. Some code has been added to the store_xxx() functions to reduce the amount
6813 of flapping under certain conditions.
6815 8. log_incoming_interface used to affect only the <= reception log lines. Now
6816 it causes the local interface and port to be added to several more SMTP log
6817 lines, for example "SMTP connection from", and rejection lines.
6819 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
6821 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
6823 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
6824 could be overwritten at the end of the current message (or the start of a
6825 new message if it was set in a HELO ACL). The value is now preserved for
6826 the duration of the SMTP connection.
6828 12. If a transport had a headers_rewrite setting, and a matching header line
6829 contained an unqualified address, that address was qualified, even if it
6830 did not match any rewriting rules. The underlying bug was that the values
6831 of the flags that permit the existence of unqualified sender and recipient
6832 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
6833 non-local messages, and by -bnq for local messages) were not being
6834 preserved with the message after it was received.
6836 13. When Exim was logging an SMTP synchronization error, it could sometimes log
6837 "next input=" as part of the text comprising the host identity instead of
6838 the correct text. The code was using the same buffer for two different
6839 strings. However, depending on which order the printing function evaluated
6840 its arguments, the bug did not always show up. Under Linux, for example, my
6841 test suite worked just fine.
6843 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
6844 doesn't work with some older versions of Perl. It has been changed to "my",
6845 which in any case is probably the better facility to use.
6847 15. A really picky compiler found some instances of statements for creating
6848 error messages that either had too many or two few arguments for the format
6851 16. The size of the buffer for calls to the DNS resolver has been increased
6852 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
6853 for addresses that have a lot of PTR records. This alleviates a problem; it
6854 does not fully solve it.
6856 17. A dnsdb lookup for PTR records that receives more data than will fit in the
6857 buffer now truncates the list and logs the incident, which is the same
6858 action as happens when Exim is looking up a host name and its aliases.
6859 Previously in this situation something unpredictable would happen;
6860 sometimes it was "internal error: store_reset failed".
6862 18. If a server dropped the connection unexpectedly when an Exim client was
6863 using GnuTLS and trying to read a response, the client delivery process
6864 crashed while trying to generate an error log message.
6866 19. If a "warn" verb in an ACL added multiple headers to a message in a single
6867 string, for example:
6869 warn message = H1: something\nH2: something
6871 the text was added as a single header line from Exim's point of view
6872 though it ended up OK in the delivered message. However, searching for the
6873 second and subsequent header lines using $h_h2: did not work. This has been
6874 fixed. Similarly, if a system filter added multiple headers in this way,
6875 the routers could not see them.
6877 20. Expanded the error message when iplsearch is called with an invalid key to
6878 suggest using net-iplsearch in a host list.
6880 21. When running tests using -bh, any delays imposed by "delay" modifiers in
6881 ACLs are no longer actually imposed (and a message to that effect is
6884 22. If a "gecos" field in a passwd entry contained escaped characters, in
6885 particular, if it contained a \" sequence, Exim got it wrong when building
6886 a From: or a Sender: header from that name. A second bug also caused
6887 incorrect handling when an unquoted " was present following a character
6888 that needed quoting.
6890 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
6891 was not being matched caselessly.
6893 24. Arranged for all hyphens in the exim.8 source to be escaped with
6896 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
6897 properly. Recipient callout cache records were still being keyed to include
6898 the sender, even when use_sender was set false. This led to far more
6899 callouts that were necessary. The sender is no longer included in the key
6900 when use_sender is false.
6902 26. Added "control = submission" modifier to ACLs.
6904 27. Added the ${base62d: operator to decode base 62 numbers.
6906 28. dnsdb lookups can now access SRV records.
6908 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
6909 the configuration file.
6911 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
6912 (-v) mode. This makes the output for a verbose queue run more intelligible.
6914 31. Added a use_postmaster feature to recipient callouts.
6916 32. Added the $body_zerocount variable, containing the number of binary zero
6917 bytes in the message body.
6919 33. The time of last modification of the "new" subdirectory is now used as the
6920 "mailbox time last read" when there is a quota error for a maildir
6923 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
6925 35. Added +ignore_unknown as a special item in host lists.
6927 36. Code for decoding IPv6 addresses in host lists is now included, even if
6928 IPv6 support is not being compiled. This fixes a bug in which an IPv6
6929 address was recognized as an IP address, but was then not correctly decoded
6930 into binary, causing unexpected and incorrect effects when compared with
6937 1. Very minor rewording of debugging text in manualroute to say "list of
6938 hosts" instead of "hostlist".
6940 2. If verify=header_syntax was set, and a header line with an unqualified
6941 address (no domain) and a large number of spaces between the end of the
6942 name and the colon was received, the reception process suffered a buffer
6943 overflow, and (when I tested it) crashed. This was caused by some obsolete
6944 code that should have been removed. The fix is to remove it!
6946 3. When running in the test harness, delay a bit after writing a bounce
6947 message to get a bit more predictability in the log output.
6949 4. Added a call to search_tidyup() just before forking a reception process. In
6950 theory, someone could use a lookup in the expansion of smtp_accept_max_
6951 per_host which, without the tidyup, could leave open a database connection.
6953 5. Added the variables $recipient_data and $sender_data which get set from a
6954 lookup success in an ACL "recipients" or "senders" condition, or a router
6955 "senders" option, similar to $domain_data and $local_part_data.
6957 6. Moved the writing of debug_print from before to after the "senders" test
6960 7. Change 4.31/66 (moving the time when the Received: is generated) caused
6961 problems for message scanning, either using a data ACL, or using
6962 local_scan() because the Received: header was not generated till after they
6963 were called (in order to set the time as the time of reception completion).
6964 I have revised the way this works. The header is now generated after the
6965 body is received, but before the ACL or local_scan() are called. After they
6966 are run, the timestamp in the header is updated.
6972 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
6973 before starting a queue runner without re-exec. This happened only when
6974 deliver_drop_privilege was set or when the Exim user was set to root. The
6975 effect of the bug was that timeouts during subsequent deliveries caused
6976 crashes instead of being properly handled. The handler is now left at its
6977 default (and expected) setting.
6979 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
6980 message, again when deliver_drop_privilege is set or Exim is run as root.
6981 The bug described in (1) was not present in this case, but the tidying up
6982 of the other signals was missing. I have made the two cases consistent.
6984 3. The ignore_target_hosts setting on a manualroute router was being ignored
6985 for hosts that were looked up using the /MX notation.
6987 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
6990 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
6991 operated on the sender address. After changing the $sender_address to <>
6992 for the sender address verify, Exim was re-instated it as the original
6993 (before rewriting) address, but remembering that it had rewritten it, so it
6994 wasn't rewriting it again. This bug also had the effect of breaking the
6995 sender address verification caching when the sender address was rewritten.
6997 6. The ignore_target_hosts option was being ignored by the ipliteral router.
6998 This has been changed so that if the ip literal address matches
6999 ignore_target_hosts, the router declines.
7001 7. Added expansion conditions match_domain, match_address, and match_local_
7002 part (NOT match_host).
7004 8. The placeholder for the Received: header didn't have a length field set.
7006 9. Added code to Exim itself and to exim_lock to test for a specific race
7007 condition that could lead to file corruption when using MBX delivery. The
7008 issue is with the lockfile that is created in /tmp. If this file is removed
7009 after a process has opened it but before that process has acquired a lock,
7010 there is the potential for a second process to recreate the file and also
7011 acquire a lock. This could lead to two Exim processes writing to the file
7012 at the same time. The added code performs the same test as UW imapd; it
7013 checks after acquiring the lock that its file descriptor still refers to
7014 the same named file.
7016 10. The buffer for building added header lines was of fixed size, 8192 bytes.
7017 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
7020 11. Added the smtp_active_hostname option. If used, this will typically be made
7021 to depend on the incoming interface address. Because $interface_address is
7022 not set up until the daemon has forked a reception process, error responses
7023 that can happen earlier (such as "too many connections") no longer contain
7026 12. If an expansion in a condition on a "warn" statement fails because a lookup
7027 defers, the "warn" statement is abandoned, and the next ACL statement is
7028 processed. Previously this caused the whole ACL to be aborted.
7030 13. Added the iplsearch lookup type.
7032 14. Added ident_timeout as a log selector.
7034 15. Added tls_certificate_verified as a log selector.
7036 16. Added a global option tls_require_ciphers (compare the smtp transport
7037 option of the same name). This controls incoming TLS connections.
7039 17. I finally figured out how to make tls_require_ciphers do a similar thing
7040 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
7041 before starting the TLS session.
7043 18. Tabs are now shown as \t in -bP output.
7045 19. If the log selector return_path_on_delivery was set, Exim crashed when
7046 bouncing a message because it had too many Received: header lines.
7048 20. If two routers both had headers_remove settings, and the first one included
7049 a superfluous trailing colon, the final name in the first list and the
7050 first name in the second list were incorrectly joined into one item (with a
7051 colon in the middle).
7057 1. Added -C and -D options to the exinext utility, mainly to make it easier
7058 to include in the automated testing, but these could be helpful when
7059 multiple configurations are in use.
7061 2. The exinext utility was not formatting the output nicely when there was
7062 an alternate port involved in the retry record key, nor when there was a
7063 message id as well (for retries that were specific to a specific message
7064 and a specific host). It was also confused by IPv6 addresses, because of
7065 the additional colons they contain. I have fixed the IPv4 problem, and
7066 patched it up to do a reasonable job for IPv6.
7068 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
7069 delivery, the log line now contains "pipelined" if PIPELINING was used.
7071 4. An SMTP transport process used to panic and die if the bind() call to set
7072 an explicit outgoing interface failed. This has been changed; it is now
7073 treated in the same way as a connect() failure.
7075 5. A reference to $sender_host_name in the part of a conditional expansion
7076 that was being skipped was still causing a DNS lookup. This no longer
7079 6. The def: expansion condition was not recognizing references to header lines
7080 that used bh_ and bheader_.
7082 7. Added the _cache feature to named lists.
7084 8. The code for checking quota_filecount in the appendfile transport was
7085 allowing one more file than it should have been.
7087 9. For compatibility with Sendmail, the command line option
7095 and sets the incoming protocol and host name (for trusted callers). The
7096 host name and its colon can be omitted when only the protocol is to be set.
7097 Note the Exim already has two private options, -pd and -ps, that refer to
7098 embedded Perl. It is therefore impossible to set a protocol value of "d" or
7099 "s", but I don't think that's a major issue.
7101 10. A number of refactoring changes to the code, none of which should affect
7104 (a) The number of logging options was getting close to filling up the
7105 32-bit word that was used as a bit map. I have split them into two classes:
7106 those that are passed in the argument to log_write(), and those that are
7107 only ever tested independently outside of that function. These are now in
7108 separate 32-bit words, so there is plenty of room for expansion again.
7109 There is no change in the user interface or the logging behaviour.
7111 (b) When building, for example, log lines, the code previously used a
7112 macro that called string_cat() twice, in order to add two strings. This is
7113 not really sufficiently general. Furthermore, there was one instance where
7114 it was actually wrong because one of the argument was used twice, and in
7115 one call a function was used. (As it happened, calling the function twice
7116 did not affect the overall behaviour.) The macro has been replaced by a
7117 function that can join an arbitrary number of extra strings onto a growing
7120 (c) The code for expansion conditions now uses a table and a binary chop
7121 instead of a serial search (which was left over from when there were very
7122 few conditions). Also, it now recognizes conditions like "pam" even when
7123 the relevant support is not compiled in: a suitably worded error message is
7124 given if an attempt is made to use such a condition.
7126 11. Added ${time_interval:xxxxx}.
7128 12. A bug was causing one of the ddress fields not to be passed back correctly
7129 from remote delivery subprocesses. The field in question was not being
7130 subsequently used, so this caused to problems in practice.
7132 13. Added new log selectors queue_time and deliver_time.
7134 14. Might have fixed a bug in maildirsizefile handling that threw up
7135 "unexpected character" debug warnings, and recalculated the data
7136 unnecessarily. In any case, I expanded the warning message to give more
7139 15. Added the message "Restricted characters in address" to the statements in
7140 the default ACL that block characters like @ and % in local parts.
7142 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
7143 Three changes have been made:
7145 (a) There was a serious bug; a negative response to MAIL caused the whole
7146 recipient domain to be cached as invalid, thereby blocking all messages
7147 to all local parts at the same domain, from all senders. This bug has
7148 been fixed. The domain is no longer cached after a negative response to
7149 MAIL if the sender used is not empty.
7151 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
7154 (c) A new callout option, "use_sender" has been added for people who want
7155 the modified behaviour.
7161 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
7164 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
7165 indeed breaks things for older releases.
7167 3. Added additional logging to the case where there is a problem reading data
7168 from a filter that is running in a subprocess using a pipe, in order to
7169 try to track down a specific problem.
7171 4. Testing facility fudge: when running in the test harness and attempting
7172 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
7173 getting "No route to host". Convert this to a timeout.
7175 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
7178 6. Some OS don't have socklen_t but use size_t instead. This affects the
7179 fifth argument of getsockopt() amongst other things. This is now
7180 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
7181 can be set for individual OS. I have set it for SunOS5, OSF1, and
7182 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
7183 some earlier ones do not.
7185 7. Change 4.30/15 was not doing the test caselessly.
7187 8. The standard form for an IPv6 address literal was being rejected by address
7188 parsing in, for example, MAIL and RCPT commands. An example of this kind of
7189 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7190 this, as well as the form without the "IPv6" on the front (but only when
7191 address literals are enabled, of course).
7193 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
7195 10. Exim crashed if a message with an empty sender address specified by -f
7196 encountered a router with an errors_to setting. This could be provoked only
7197 by a command such as
7201 where an empty string was supplied; "<>" did not hit this bug.
7203 11. Installed PCRE release 4.5.
7205 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
7206 remained set. It is now erased.
7208 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
7209 times from message ids (which are base 36 rather than the normal 62).
7211 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
7212 were being counted as actual protocol errors, and logged if the log
7213 selector +smtp_protocol_error was set. One cannot be perfect in this test,
7214 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
7215 and DATA following a set of rejected RCPTs do not count as protocol errors.
7216 In other words, Exim assumes they were pipelined, though this may not
7217 actually be the case. Of course, in all cases the client gets an
7218 appropriate error code.
7220 15. If a lookup fails in an ACL condition, a message about the failure may
7221 be available; it is used if testing the ACL cannot continue, because most
7222 such messages specify what the cause of the deferral is. However, some
7223 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
7224 that caused an old message to be retained and used if a later statement
7225 caused a defer, replacing the real cause of the deferral.
7227 16. If an IP address had so many PTR records that the DNS lookup buffer
7228 was not large enough to hold them, Exim could crash while trying to process
7229 the truncated data. It now detects and logs this case.
7231 17. Further to 4.21/58, another change has been made: if (and only if) the
7232 first line of a message (the first header line) ends with CRLF, a bare LF
7233 in a subsequent header line has a space inserted after it, so as not to
7234 terminate the header.
7236 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
7237 unnecessarily, used atoi() instead of strtol(), and didn't check the
7238 termination when getting file sizes from file names by regex.
7240 19. Completely re-implemented the support for maildirsize files, in the light
7241 of a number of problems with the previous contributed implementation
7242 (4.30/29). In particular:
7244 . If the quota is zero, the maildirsize file is maintained, but no quota is
7247 . If the maildir directory does not exist, it is created before any attempt
7248 to write a maildirsize file.
7250 . The quota value in the file is just a cache; if the quota is changed in
7251 the transport, the new value overrides.
7253 . A regular expression is available for excluding directories from the
7256 20. The autoreply transport checks the characters in options that define the
7257 message's headers; it allows continued headers, but it was checking with
7258 isspace() after an embedded newline instead of explicitly looking for a
7261 21. If all the "regular" hosts to which an address was routed had passed their
7262 expiry times, and had not reached their retry times, the address was
7263 bounced, even if fallback hosts were defined. Now Exim should go on to try
7266 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
7267 equivalent code in the SMTP transport. Some hosts send humungous responses
7268 to HELO/EHLO, more than 1024 it seems.
7270 23. Refactoring: code in filter.c used (void *) for "any old type" but this
7271 gives compiler warnings in some environments. I've now done it "properly",
7274 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
7275 (because of problems with the built-in one) was declared to return uschar *
7276 instead of char *, causing compiler failure.
7278 25. Fixed a file descriptor leak when processing alias/forward files.
7280 26. Fixed a minor format string issue in dbfn.c.
7282 27. Typo in exim.c: ("dmbnz" for "dbmnz").
7284 28. If a filter file refered to $h_xxx or $message_headers, and the headers
7285 contained RFC 2047 "words", Exim's memory could, under certain conditions,
7288 29. When a sender address is verified, it is cached, to save repeating the test
7289 when there is more than one recipient in a message. However, when the
7290 verification involves a callout, it is possible for different callout
7291 options to be set for different recipients. It is too complicated to keep
7292 track of this in the cache, so now Exim always runs a verification when a
7293 callout is required, relying on the callout cache for the optimization.
7294 The overhead is duplication of the address routing, but this should not be
7297 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
7298 to be verified with callout=postmaster, and the main callout worked but the
7299 postmaster check failed, the verification correctly failed. However, if a
7300 subsequent RCPT command asked for sender verification *without* the
7301 postmaster check, incorrect caching caused this verification also to fail,
7304 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
7305 it was not caching the DNS options (qualify_single, search_parents) that
7306 were used when the lookup failed. A subsequent lookup with different
7307 options therefore always gave the same answer, though there were cases
7308 where it should not have. (Example: a "domains = !$mx_any" option on a
7309 dnslookup router: the "domains" option is always processed without any
7310 widening, but the router might have qualify_single set.) Now Exim uses the
7311 cached value only when the same options are set.
7313 32. Added John Jetmore's "exipick" utility to the distribution.
7315 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
7316 than a timeout (e.g. a certificate is required, and is not provided), an
7317 Exim server now closes the connection immediately. Previously it waited for
7318 the client to close - but if the client is SSL, it seems that they each
7319 wait for each other, leading to a delay before one of them times out.
7321 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
7322 maintained 0.8.x compatibility because I don't think many are using it, and
7323 it is clearly obsolete.
7325 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
7328 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
7329 client certificate was expired. A simple patch fixes this, though I don't
7330 understand the full logic of why the verify callback is called multiple
7333 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
7334 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
7335 which causes problems with some clients (such as the Certicom SSL Plus
7336 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
7337 disables the coutermeasure allowing Eudora to connect."
7339 38. Exim was not checking that a write() to a log file succeeded. This could
7340 lead to Bad Things if a log got too big, in particular if it hit a file
7341 size limit. Exim now panics and dies if it cannot write to a log file, just
7342 as it does if it cannot open a log file.
7344 39. Modified OS/Makefile-Linux so that it now contains
7346 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
7348 The two -D definitions ensure that Exim is compiled with large file
7349 support, which makes it possible to handle log files that are bigger than
7352 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
7353 instance) a domain was checked against a named list that involved a lookup,
7354 causing $domain_data to be set, then another domain was checked against the
7355 same list, then the first domain was re-checked, the value of $domain_data
7356 after the final check could be wrong. In particular, if the second check
7357 failed, it could be set empty. This bug probably also applied to
7360 41. The strip_trailing_dot option was not being applied to the address given
7361 with the -f command-line option.
7363 42. The code for reading a message's header from the spool was incrementing
7364 $received_count, but never initializing it. This meant that the value was
7365 incorrect (doubled) while delivering a message in the same process in which
7366 it was received. In the most common configuration of Exim, this never
7367 happens - a fresh exec is done - but it can happen when
7368 deliver_drop_privilege is set.
7370 43. When Exim logs an SMTP synchronization error - client data sent too soon -
7371 it now includes up to 150 characters of the unexpected data in the log
7374 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
7375 and building data strings. The size of both of these buffers was 10 000
7376 bytes - far larger than anybody would *ever* want, thought I. Needless to
7377 say, somebody hit the limit. I have increased the maximum line length to
7378 20 000 and the maximum data length of concatenated lines to 100 000. I have
7379 also fixed two bugs, because there was no checking on these buffers. Tsk,
7380 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
7381 buffer is too small.
7383 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
7384 lsearch lookups. Now it does.
7386 46. When parsing a route_list item in a manualroute router, a fixed-length
7387 buffer was used for the list of hosts. I made this 1024 bytes long,
7388 thinking that nobody would ever have a list of hosts that long. Wrong.
7389 Somebody had a whole pile of complicated expansion conditions, and the
7390 string was silently truncated, leading to an expansion error. It turns out
7391 that it is easier to change to an unlimited length (owing to other changes
7392 that have happened since this code was originally written) than to build
7393 structure for giving a limitation error. The length of the item that
7394 expands into the list of hosts is now unlimited.
7396 47. The lsearch lookup could not handle data where the length of text line was
7397 more than 4095 characters. Such lines were truncated, leading to shortened
7398 data being returned. It should now handle lines of any length.
7400 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
7401 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
7404 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
7405 replacing strings such as BIN_DIRECTORY in a source file: the replacement
7406 no longer happens in comment lines. A list of replacements is now placed
7407 at the head of all of the source files, except those whose only change is
7408 to replace PERL_COMMAND in the very first #! line.
7410 50. Replaced the slow insertion sort in queue.c, for sorting the list of
7411 messages on the queue, with a bottom-up merge sort, using code contributed
7412 by Michael Haardt. This should make operations like -bp somewhat faster on
7413 large queues. It won't affect queue runners, except when queue_run_in_order
7416 51. Installed eximstats 1.31 in the distribution.
7418 52. Added support for SRV lookups to the dnslookup router.
7420 53. If an ACL referred to $message_body or $message_body_end, the value was not
7421 reset for any messages that followed in the same SMTP session.
7423 54. The store-handling optimization for building very long strings was not
7424 differentiating between the different store pools. I don't think this
7425 actually made any difference in practice, but I've tidied it.
7427 55. While running the routers to verify a sender address, $sender_address
7428 was still set to the sender address. This is wrong, because when routing to
7429 send a bounce to the sender, it would be empty. Therefore, I have changed
7430 it so that, while verifying a sender address, $sender_address is set to <>.
7431 (There is no change to what happens when verifying a recipient address.)
7433 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
7434 target A or AAAA records (if not already returned) without resetting the
7435 qualify_single or search_parents options of the DNS resolver. These are
7436 inappropriate in this case because the targets of MX and SRV records must
7437 be FQDNs. A broken DNS record could cause trouble if it happened to have a
7438 target that, when qualified, matched something in the local domain. These
7439 two options are now turned off when doing these lookups.
7441 57. It seems that at least some releases of Reiserfs (which does not have the
7442 concept of a fixed number of inodes) returns zero and not -1 for the
7443 number of available inodes. This interacted badly with check_spool_inodes,
7444 which assumed that -1 was the "no such thing" setting. What I have done is
7445 to check that the total number of inodes is greater than zero before doing
7446 the test of how many are available.
7448 58. When a "warn" ACL statement has a log_message modifier, the message is
7449 remembered, and not repeated. This is to avoid a lot of repetition when a
7450 message has many recipients that cause the same warning to be written.
7451 However, Exim was preserving the list of already written lines for an
7452 entire SMTP session, which doesn't seem right. The memory is now reset if a
7453 new message is started.
7455 59. The "rewrite" debugging flag was not showing the result of rewriting in the
7456 debugging output unless log_rewrite was also set.
7458 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
7459 of (int)(handle) when we know that handle contains (void *)(-1).
7461 61. The Exim daemon panic-logs an error return when it closes the incoming
7462 connection. However "connection reset by peer" seems to be common, and
7463 isn't really an error worthy of noting specially, so that particular error
7466 62. When Exim is trying to find all the local interfaces, it used to panic and
7467 die if the ioctl to get the interface flags failed. However, it seems that
7468 on at least one OS (Solaris 9) it is possible to have an interface that is
7469 included in the list of interfaces, but for which you get a failure error
7470 for this call. This happens when the interface is not "plumbed" into a
7471 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
7472 failure of the "get flags" call assumes that the interface is down.
7474 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
7475 makes life easier for people who are doing arithmetic on fields extracted
7476 from dates, where you often get leading zeros that should not be
7477 interpreted as octal.
7479 64. Added qualify_domain to the redirect router, to override the global
7482 65. If a pathologically long header line contained very many addresses (the
7483 report of this problem mentioned 10 000) and each of them was rewritten,
7484 Exim could use up a very large amount of memory. (It kept on making new
7485 copies of the header line as it rewrote, and never released the old ones.)
7486 At the expense of a bit more processing, the header rewriting function has
7487 been changed so that it no longer eats memory in this way.
7489 66. The generation of the Received: header has been moved from the time that a
7490 message starts to be received, to the time that it finishes. The timestamp
7491 in the Received: header should now be very close to that of the <= log
7492 line. There are two side-effects of this change:
7494 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
7495 logged header lines no longer include the local Received: line, because
7496 it has not yet been created. The same applies to a copy of the message
7497 that is returned to a non-SMTP sender when a message is rejected.
7499 (b) When a filter file is tested using -bf, no additional Received: header
7500 is added to the test message. After some thought, I decided that this
7503 This change does not affect the value of $received_for. It is still set
7504 after address rewriting, but before local_scan() is called.
7506 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
7508 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
7509 gave an unhelpful panic error message, and a defer error. I have managed to
7510 change this behaviour so that it now rejects any supplied certificate,
7511 which seems right, as the list of acceptable certificates is empty.
7513 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
7514 gave an unhelpful defer error. I have not managed to make this reject any
7515 supplied certificates, but the error message it gives is "no certificate
7516 supplied", which is not helpful.
7518 70. exigrep's output now also includes lines that are not associated with any
7519 message, but which match the given pattern. Implemented by a patch from
7520 Martin Sluka, which also tidied up the Perl a bit.
7522 71. Recipient callout verification, like sender verification, was using <> in
7523 the MAIL FROM command. This isn't really the right thing, since the actual
7524 sender may affect whether the remote host accepts the recipient or not. I
7525 have changed it to use the actual sender in the callout; this means that
7526 the cache record is now keyed on a recipient/sender pair, not just the
7527 recipient address. There doesn't seem to be a real danger of callout loops,
7528 since a callout by the remote host to check the sender would use <>.
7529 [SEE ABOVE: changed after hitting problems.]
7531 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
7532 temporary errors. However, in the case of such a code being given after
7533 the end of a data transmission (i.e. after ".") Exim was failing to write
7534 a retry record for the message. (Yes, there was some broken host that was
7535 actually sending 8xx at this point.)
7537 73. An unknown lookup type in a host list could cause Exim to panic-die when
7538 the list was checked. (An example that provoked this was putting <; in the
7539 middle of a list instead of at the start.) If this happened during a DATA
7540 ACL check, a -D file could be left lying around. This kind of configuration
7541 error no longer causes Exim to die; instead it causes a defer error. The
7542 incident is still logged to the main and panic logs.
7544 74. Buglet left over from Exim 3 conversion. The message "too many messages
7545 in one connection" was written to the rejectlog but not the mainlog, except
7546 when address rewriting (yes!) was being logged.
7548 75. Added write_rejectlog option.
7550 76. When a system filter was run not as root (that is, when system_filter_user
7551 was set), the values of the $n variables were not being returned to the
7552 main process; thus, they were not subsequently available in the $sn
7555 77. Added +return_path_on_delivery log selector.
7557 78. A connection timeout was being treated differently from recipients deferred
7558 when testing hosts_max_try with a message that was older than the host's
7559 retry timeout. (The host should not be counted, thus allowing all hosts to
7560 be tried at least once before bouncing.) This may have been the cause of an
7561 occasionally reported bug whereby a message would remain on the queue
7562 longer than the retry timeout, but would be bounced if a delivery was
7563 forced. I say "may" because I never totally pinned down the problem;
7564 setting up timeout/retry tests is difficult. See also the next item.
7566 79. The ultimate address timeout was not being applied to errors that involved
7567 a combination of host plus message (for example, a timeout on a MAIL
7568 command). When an address resolved to a number of possible hosts, and they
7569 were not all tried for each delivery (e.g. because of hosts_max_try), a
7570 message could remain on the queue longer than the retry timeout.
7572 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
7575 81. Fixed an obscure SMTP outgoing bug which required at least the following
7576 conditions: (a) there was another message waiting for the same server;
7577 (b) the server returned 5xx to all RCPT commands in the first message so
7578 that the message was not completed; (c) the server dropped the connection
7579 or gave a negative response to the RSET that Exim sends to abort the
7580 transaction. The observed case was a dropped connection after DATA that had
7581 been sent in pipelining mode. That is, the server had advertised PIPELINING
7582 but was not implementing it correctly. The effect of the bug was incorrect
7583 behaviour, such as trying another host, and this could lead to a crash.
7589 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
7590 and daemon.c were passed as pointers to ints; they should have been
7591 pointers to socklen_t variables (which are typically unsigned ints).
7593 2. Some signed/unsigned type warnings in the os.c file for Linux have been
7596 3. Fixed a really odd bug that affected only the testing scheme; patching a
7597 certain fixed string in the binary changed the value of another string that
7598 happened to be identical to the end of the original first string.
7600 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
7601 name", it returns that address as the IP address. On some operating
7602 systems (e.g. Solaris), it also passes back the IP address string as the
7603 "host name". However, on others (e.g. Linux), it passes back an empty
7604 string. Exim wasn't checking for this, and was changing the host name to an
7605 empty string, assuming it had been canonicalized.
7607 5. Although rare, it is permitted to have more than one PTR record for a given
7608 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
7609 all the names associated with an address, because they do in Solaris.
7610 However, it seems that they do not in Linux for data that comes from the
7611 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
7612 I found this out when I moved to a new Linux workstation and tried to run
7613 the Exim test suite.
7615 To get round this problem I have changed the code so that it now does its
7616 own call to the DNS to look up PTR records when searching for a host name.
7617 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
7618 addresses that are only in /etc/hosts are still found.
7620 This behaviour is, however, controlled by an option called host_lookup_
7621 order, which defaults to "bydns:byaddr". If people want to use the other
7622 order, or indeed, just use one or the other means of lookup, they can
7623 specify it in this variable.
7625 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
7626 some operating systems, this comes back from gethostbyaddr() as an empty
7627 string, and this is what Exim used to test for. However, it seems that in
7628 other systems, "." is yielded. Exim now tests for this case too.
7630 7. The values of check_spool_space and check_log_space are now held internally
7631 as a number of kilobytes instead of an absolute number of bytes. If a
7632 numbers is specified without 'K' or 'M', it is rounded up to the nearest
7633 kilobyte. This means that much larger values can be stored.
7635 8. Exim monitor: an attempt to get the action menu when not actually pointing
7636 at a message produces an empty menu entitled "No message selected". This
7637 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
7638 no entries in it ("Shell widget menu has zero width and/or height"). So I
7639 have added a single, blank menu entry in this case.
7641 9. Added ${quote_local_part.
7643 10. MIME decoding is now applied to the contents of Subject: header lines when
7646 11. Now that a reference to $sender_host_address automatically causes a reverse
7647 lookup to occur if necessary (4.13/18), there is no need to arrange for a
7648 host lookup before query-style lookups in lists that might use this
7649 variable. This has therefore been abolished, and the "net-" prefix is no
7650 longer necessary for query-style lookups.
7652 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
7653 have been a typo for LFLAGS, so it has been changed.
7655 13. The install script calls Exim with "-C /dev/null" in order to find the
7656 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
7657 to be output. However, since Exim outputs its version number before the
7658 error, it didn't break the script. It just looked ugly. I fixed this by
7659 always allowing "-C /dev/null" if the caller is root.
7661 14. Ignore overlarge ACL variable number when reading spool file - insurance
7662 against a later release with more variables having written the file.
7664 15. The standard form for an IPv6 address literal was being rejected by EHLO.
7665 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
7666 this, as well as the form without the "IPv6" on the front.
7668 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
7669 OS/Makefile-Darwin file.
7671 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
7672 only to LDAP libraries that do not have LDAP_OPT_DEREF.
7674 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
7675 variable. However, some OS use ints for inodes. I've added cast to long int
7676 to get rid of the compiler warning.
7678 19. I had forgotten to lock out "/../" in configuration file names when
7679 ALT_CONFIG_PREFIX was set.
7681 20. Routers used for verification do not need to specify transports. However,
7682 if such a router generated a host list, and callout was configured, Exim
7683 crashed, because it could not find a port number from the (non-existent)
7684 transport. It now assumes port 25 in this circumstance.
7686 21. Added the -t option to exigrep.
7688 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
7689 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
7690 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
7691 them is set, LOOKUP_LSEARCH is forced.
7693 23. "exim -bV" now outputs a list of lookups that are included in the binary.
7695 24. Added sender and host information to the "rejected by local_scan()" log
7696 line; previously there was no indication of these.
7698 25. Added .include_if_exists.
7700 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
7701 Linux. It turns out that not all file systems support this. Apparently some
7702 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
7703 do it.) To cope with this, the error EINVAL, which means that sync-ing is
7704 not supported on the file descriptor, is now ignored when Exim is trying to
7705 sync a directory. This applies only to Linux.
7707 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
7709 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
7710 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
7713 29. Added support for maildirsize files from supplied patch (modified a bit).
7715 30. The use of :fail: followed by an empty string could lead Exim to respond to
7716 sender verification failures with (e.g.):
7718 550 Verification failed for <xxx>
7719 550 Sender verify failed
7721 where the first response line was missing the '-' that indicates it is not
7722 the final line of the response.
7724 31. The loop for finding the name of the user that called Exim had a hardwired
7725 limit of 10; it now uses the value of finduser_retries, which is used for
7726 all other user lookups.
7728 32. Added $received_count variable, available in data and not_smtp ACLs, and at
7731 33. Exim was neglecting to zero errno before one call of strtol() when
7732 expanding a string and expecting an integer value. On some systems this
7733 resulted in spurious "integer overflow" errors. Also, it was casting the
7734 result into an int without checking.
7736 34. Testing for a connection timeout using "timeout_connect" in the retry rules
7737 did not work. The code looks as if it has *never* worked, though it appears
7738 to have been documented since at least release 1.62. I have made it work.
7740 35. The "timeout_DNS" error in retry rules, also documented since at least
7741 1.62, also never worked. As it isn't clear exactly what this means, and
7742 clearly it isn't a major issue, I have abolished the feature by treating it
7743 as "timeout", and writing a warning to the main and panic logs.
7745 36. The display of retry rules for -brt wasn't always showing the error code
7748 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
7749 timeout_connect_A, timeout_connect_MX.
7751 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
7752 to the empty sender.
7754 39. The daemon was not analysing the content of -oX till after it had closed
7755 stderr and disconnected from the controlling terminal. This meant that any
7756 syntax errors were only noted on the panic log, and the return code from
7757 the command was 0. By re-arranging the code a little, I've made the
7758 decoding happen first, so such errors now appear on stderr, and the return
7759 code is 1. However, the actual setting up of the sockets still happens in
7760 the disconnected process, so errors there are still only recorded on the
7763 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
7764 connections (as happens on some IP stacks) was logged at start up time as
7765 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
7766 it from "IPv6 and IPv4", which means that two separate sockets are being
7769 41. The debug output for gethostbyname2() or getipnodebyname() failures now
7770 says whether AF_INET or AF_INET6 was passed as an argument.
7772 42. Exiwhat output was messed up when time zones were included in log
7775 43. Exiwhat now gives more information about the daemon's listening ports,
7776 and whether -tls-on-connect was used.
7778 44. The "port" option of the smtp transport is now expanded.
7780 45. A "message" modifier in a "warn" statement in a non-message ACL was being
7781 silently ignored. Now an error message is written to the main and panic
7784 46. There's a new ACL modifier called "logwrite" which writes to a log file
7785 as soon as it is encountered.
7787 47. Added $local_user_uid and $local_user_gid at routing time.
7789 48. Exim crashed when trying to verify a sender address that was being
7792 49. Exim was recognizing only a space character after ".include". It now also
7793 recognizes a tab character.
7795 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
7796 extracting the relevant information from the specification. The man page no
7797 longer contains scrambled data for the -d option, and I've added a section
7798 at the front about calling Exim under different names.
7800 51. Added "extra_headers" argument to the "mail" command in filter files.
7802 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
7805 53. Installed eximstats 1.29.
7807 54. Added transport_filter_timeout as a generic transport option.
7809 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
7810 Cc: header lines. This was required by RFC 822, but it not required by RFC
7813 56. Exim used to add From:, Date:, and Message-Id: header lines to any
7814 incoming messages that did not have them. Now it does so only if the
7815 message originates locally, that is, if there is no associated remote host
7816 address. When Resent- header lines are present, this applies to the Resent-
7817 lines rather than the non-Resent- lines.
7819 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
7820 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
7822 58. Messages for configuration errors now include the name of the main
7823 configuration file - useful now that there may be more than one file in a
7824 list (.included file names were always shown).
7826 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
7827 for those rare installations that do not start the daemon as root or run it
7828 setuid root. I've cut out the call to initgroups() if the daemon is not
7831 60. The Exim user and group can now be bound into the binary as text strings
7832 that are looked up at the start of Exim's processing.
7834 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
7836 62. Added $mailstore_basename variable.
7838 63. Installed patch to sieve.c from Michael Haardt.
7840 64. When Exim failed to open the panic log after failing to open the main log,
7841 the original message it was trying to log was written to stderr and debug
7842 output, but if they were not available (the usual case in production), it
7843 was lost. Now it is written to syslog before the two lines that record the
7844 failures to open the logs.
7846 65. Users' Exim filters run in subprocesses under the user's uid. It is
7847 possible for a "deliver" command or an alias in a "personal" command to
7848 provoke an address rewrite. If logging of address rewriting is configured,
7849 this fails because the process is not running as root or exim. There may be
7850 a better way of dealing with this, but for the moment (because 4.30 needs
7851 to be released), I have disabled address rewrite logging when running a
7852 filter in a non-root, non-exim process.
7858 1. The buildconfig auxiliary program wasn't quoting the value set for
7859 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
7860 not defined. This bug was masked in 4.22 by the effect that was fixed in
7863 2. Some messages that were rejected after a message id was allocated were
7864 shown as "incomplete" by exigrep. It no longer does this for messages that
7865 are rejected by local_scan() or the DATA or non-SMTP ACLs.
7867 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
7868 have allow_domain_literals set, the ID did not get logged in the <= line.
7869 Domain literals are now always recognized in Message-ID: header lines.
7871 4. The first argument for a ${extract expansion item is the key name or field
7872 number. Leading and trailing spaces in this item were not being ignored,
7873 causing some misleading effects.
7875 5. When deliver_drop_privilege was set, single queue runner processes started
7876 manually (i.e. by the command "exim -q") or by the daemon (which uses the
7877 same command in the process it spins off) were not dropping privilege.
7879 6. When the daemon running as "exim" started a queue runner, it always
7880 re-executed Exim in the spun-off process. This is a waste of effort when
7881 deliver_drop_privilege is set. The new process now just calls the
7882 queue-runner function directly.
7888 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
7891 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
7892 ignored. Though the use of -oP was forcing the writing of a pid file, it
7893 was always written to the default place.
7895 3. If the message "no IP address found for host xxxx" is generated during
7896 incoming verification, it is now followed by identification of the incoming
7897 connection (so you can more easily find what provoked it).
7899 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
7901 5. Added some features to "harden" Exim a bit more against certain attacks:
7903 (a) There is now a build-time option called FIXED_NEVER_USERS that can
7904 be put in Local/Makefile. This is like the never_users runtime option,
7905 but it cannot be overridden. The default setting is "root".
7907 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
7908 prefix string with which any file named in a -C command line option
7911 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
7912 is retained for -C and -D only if the caller of Exim is root. Without
7913 it, the exim user may also use -C and -D and retain privilege.
7915 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
7916 command line option is disabled.
7918 6. Macro names set by the -D option must start with an upper case letter, just
7919 like macro names defined in the configuration file.
7921 7. Added "dereference=" facility to LDAP.
7923 8. Two instances of the typo "uknown" in the source files are fixed.
7925 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
7926 the Configure-Makefile script screwed up while processing it.
7928 10. Incorporated PCRE 4.4.
7930 11. The SMTP synchronization check was not operating right at the start of an
7931 SMTP session. For example, it could not catch a HELO sent before the client
7932 waited for the greeting. There is now a check for outstanding input at the
7933 point when the greeting is written. Because of the duplex, asynchronous
7934 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
7935 way, but not yet received, when the check is performed.
7937 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
7938 on TCP/IP sockets, because this apparently causes some broken clients to
7941 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
7942 unchanged) from the Cygwin maintainer.
7944 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
7945 was supported instead of testing for mbx. Effectively a typo.
7947 15. The spa authenticator server code was not checking that the input it
7948 received was valid base64.
7950 16. The debug output line for the "set" modifier in ACLs was not showing the
7951 name of the variable that was being set.
7953 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
7955 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
7956 lookup on its own. Something else was needed to trigger the lookup. For
7957 example, a match in host_lookup or the need for a host name in a host list.
7958 Now, if $sender_host_name is referenced and the host name has not yet been
7959 looked up, a lookup is performed. If the lookup fails, the variable remains
7960 empty, and $host_lookup_failed is set to "1".
7962 19. Added "eqi" as a case-independent comparison operator.
7964 20. The saslauthd authentication condition could segfault if neither service
7965 nor realm was specified.
7967 21. If an overflowing value such as "2048M" was set for message_size_limit, the
7968 error message that was logged was misleading, and incoming SMTP
7969 connections were dropped. The message is now more accurate, and temporary
7970 errors are given to SMTP connections.
7972 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
7973 (except RSET) with a 421 error, until QUIT is received. However, it was
7974 failing to send a response to QUIT.
7976 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
7977 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
7978 ACL is now run after the helo_try_verify_hosts code.
7980 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
7981 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
7982 case-independent, so other case variants are also recognized). Apparently
7983 some systems use these upper case variants.
7985 25. If more than two messages were waiting for the same host, and a transport
7986 filter was specified for the transport, Exim sent two messages over the
7987 same TCP/IP connection, and then failed with "socket operation on non-
7988 socket" when it tried to send the third.
7990 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
7992 27. The extern definition of crypt16() in expand.c was not being excluded when
7993 the OS had its own crypt16() function.
7995 28. Added bounce_return_body as a new option, and bounce_return_size_limit
7996 as a preferred synonym for return_size_limit, both as an option and as an
7999 29. Added LIBS=-liconv to OS/Makefile-OSF1.
8001 30. Changed the default configuration ACL to relax the local part checking rule
8002 for addresses that are not in any local domains. For these addresses,
8003 slashes and pipe symbols are allowed within local parts, but the sequence
8004 /../ is explicitly forbidden.
8006 31. SPA server authentication was not clearing the challenge buffer before
8009 32. log_message in a "warn" ACL statement was writing to the reject log as
8010 well as to the main log, which contradicts the documentation and doesn't
8011 seem right (because no rejection is happening). So I have stopped it.
8013 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
8014 However, I am unable to do any testing of this.
8016 34. Fixed an infelicity in the appendfile transport. When checking directories
8017 for a mailbox, to see if any needed to be created, it was accidentally
8018 using path names with one or more superfluous leading slashes; tracing
8019 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
8021 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
8022 contents are added to the log line that is written for every discarded
8023 recipient. (Previously a log_message setting was ignored.)
8025 36. The ${quote: operator now quotes the string if it is empty.
8027 37. The install script runs exim in order to find its version number. If for
8028 some reason other than non-existence or emptiness, which it checks, it
8029 could not run './exim', it was installing it with an empty version number,
8030 i.e. as "exim-". This error state is now caught, and the installation is
8033 38. An argument was missing from the function that creates an error message
8034 when Exim fails to connect to the socket for saslauthd authentication.
8035 This could cause Exim to crash, or give a corrupted message.
8037 39. Added isip, isip4, and isip6 to ${if conditions.
8039 40. The ACL variables $acl_xx are now saved with the message, and can be
8040 accessed later in routers, transports, and filters.
8042 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
8043 strings in the file are not string-expanded.
8045 42. If a MAIL command specified a SIZE value that was too large to fit into an
8046 int variable, the check against message_size_limit failed. Such values are
8047 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
8048 day this will have to be increased, but I don't think I want to be around
8049 when emails are that large.
8056 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
8057 iconv() is not standard in FreeBSD.
8059 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
8060 IPv6 enabled. The observed symptom was a segmentation fault on return from
8061 the function os_common_find_running_interfaces() in src/os.c.
8063 3. In the check_special_case() function in daemon.c I had used "errno" as an
8064 argument name, which causes warnings on some systems. This was basically a
8065 typo, since it was named "eno" in the comments!
8067 4. The code that waits for the clock to tick (at a resolution of some fraction
8068 of a second) so as to ensure message-id uniqueness was always waiting for
8069 at least one whole tick, when it could have waited for less. [This is
8070 almost certainly not relevant at current processor speeds, where it is
8071 unlikely to ever wait at all. But we try to future-proof.]
8073 5. The function that sleeps for a time interval that includes fractions of a
8074 second contained a race. It did not block SIGALRM between setting the
8075 timer, and suspending (a couple of lines later). If the interval was short
8076 and the sigsuspend() was delayed until after it had expired, the suspension
8077 never ended. On busy systems this could lead to processes getting stuck for
8080 6. Some uncommon configurations may cause a lookup to happen in a queue runner
8081 process, before it forks any delivery processes. The open lookup caching
8082 mechanism meant that the open file or database connection was passed into
8083 the delivery process. The problem was that delivery processes always tidy
8084 up cached lookup data. This could cause a problem for the next delivery
8085 process started by the queue runner, because the external queue runner
8086 process does not know about the closure. So the next delivery process
8087 still has data in the lookup cache. In the case of a file lookup, there was
8088 no problem because closing a file descriptor in a subprocess doesn't affect
8089 the parent. However, if the lookup was caching a connection to a database,
8090 the connection was closed, and the second delivery process was likely to
8091 see errors such as "PGSQL: query failed: server closed the connection
8092 unexpectedly". The problem has been fixed by closing all cached lookups
8093 in a queue runner before running a delivery process.
8095 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
8096 seem to have the "const" qualifier which it has on other OS. I've
8099 8. Change 4.21/2 was too strict. It is only if there are two authenticators
8100 *of the same type* (client or server) with the same public name that an
8101 error should be diagnosed.
8103 9. When Exim looked up a host name for an IP address, but failed to find the
8104 original IP address when looking up the host name (a safety check), it
8105 output the message "<ip address> does not match any IP for NULL", which was
8106 confusing, to say the least. The bug was that the host name should have
8107 appeared instead of "NULL".
8109 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
8110 user that is built into the binary, and the -C or -D options is used, root
8111 privilege is dropped before the configuration file is read. In addition,
8112 logging is switched to stderr instead of the normal log files. If the
8113 configuration then re-defines the Exim user, the unprivileged environment
8114 is probably not what is expected, so Exim logs a panic warning message (but
8117 However, if deliver_drop_privilege is set, the unprivileged state may well
8118 be exactly what is intended, so the warning has been cut out in that case,
8119 and Exim is allowed to try to write to its normal log files.
8125 1. smtp_return_error_details was not giving details for temporary sender
8126 or receiver verification errors.
8128 2. Diagnose a configuration error if two authenticators have the same public
8131 3. Exim used not to create the message log file for a message until the first
8132 delivery attempt. This could be confusing when incoming messages were held
8133 for policy or load reasons. The message log file is now created at the time
8134 the message is received, and an initial "Received" line is written to it.
8136 4. The automatically generated man page for command line options had a minor
8137 bug that caused no ill effects; however, a more serious problem was that
8138 the procedure for building the man page automatically didn't always
8139 operate. Consequently, release 4.20 contains an out-of-date version. This
8140 shouldn't happen again.
8142 5. When building Exim with embedded Perl support, the script that builds the
8143 Makefile was calling 'perl' to find its compile-time parameters, ignoring
8144 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
8146 6. The freeze_tell option was not being used for messages that were frozen on
8147 arrival, either by an ACL or by local_scan().
8149 7. Added the smtp_incomplete_transaction log selector.
8151 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
8152 was accepting AUTH without a new EHLO.
8154 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
8155 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
8158 10. Logging of TCP/IP connections (when configured) now happens in the main
8159 daemon process instead of the child process, so that the TCP/IP connection
8160 count is more accurate (but it can never be perfect).
8162 11. The use of "drop" in a nested ACL was not being handled correctly in the
8163 outer ACL. Now, if condition failure induced by the nested "drop" causes
8164 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
8165 or "require"), the connection is dropped.
8167 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
8168 that yield "discard" can now be used with an "accept" or a "discard" verb,
8169 but an error is generated for any others (because I can't see a useful way
8170 to define what should happen).
8172 13. When an ACL is read dynamically from a file (or anywhere else), the lines
8173 are now processed in the same way as lines in the Exim configuration file.
8174 In particular, continuation lines are supported.
8176 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
8178 15. Added -ti meaning -t -i.
8180 16. Check for letters, digits, hyphens, and dots in the names of dnslist
8181 domains, and warn by logging if others are found.
8183 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
8184 returned from GIFCONF when Exim is trying to find the list of interfaces on
8185 a host. The code in os.c has been modified to copy each ifreq to an aligned
8186 structure in all cases.
8188 Also, in some cases, the returned ifreq's were being copied to a 'struct
8189 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
8190 means the last couple of bytes of an IPv6 address could be chopped if the
8191 ifreq contained only a normal sockaddr (14 bytes storage).
8193 18. Named domain lists were not supported in the hosts_treat_as_local option.
8194 An entry such as +xxxx was not recognized, and was treated as a literal
8197 19. Ensure that header lines added by a DATA ACL are included in the reject log
8198 if the ACL subsequently rejects the message.
8200 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
8201 MD5 (which is deprecated).
8203 21. When testing a filter file using -bf, Exim was writing a message when it
8204 took the sender from a "From " line in the message, but it was not doing so
8205 when it took $return_path from a Return-Path: header line. It now does.
8207 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
8208 with a valid header line field name (a series of printing characters
8209 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
8211 23. Changed "disc" in the source to "disk" to conform to the documentation and
8212 the book and for uniformity.
8214 24. Ignore Sendmail's -Ooption=value command line item.
8216 25. When execve() failed while trying to run a command in a pipe transport,
8217 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
8218 could be confused with a return value of 69 from the command itself. This
8219 has been changed to 127, the value the shell returns if it is asked to run
8220 a non-existent command. The wording for the related log line suggests a
8221 non-existent command as the problem.
8223 26. If received_header_text expands to an empty string, do not add a Received:
8224 header line to the message. (Well, it adds a token one on the spool, but
8225 marks it "old" so that it doesn't get used or transmitted.)
8227 27. Installed eximstats 1.28 (addition of -nt option).
8229 28. There was no check for failure on the call to getsockname() in the daemon
8230 code. This can fail if there is a shortage of resources on the system, with
8231 ENOMEM, for example. A temporary error is now given on failure.
8233 29. Contrary to the C standard, it seems that in some environments, the
8234 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
8235 program. Exim now does this explicitly; it affects the formatting of
8236 timestamps using strftime().
8238 30. If exiqsumm was given junk data, it threw up some uninitialized variable
8239 complaints. I've now initialized all the variables, to avoid this.
8241 32. Header lines added by a system filter were not being "seen" during
8242 transport-time rewrites.
8244 33. The info_callback() function passed to OpenSSL is set up with type void
8245 (*)(SSL *, int, int), as described somewhere. However, when calling the
8246 function (actually a macro) that sets it up, the type void(*)() is
8247 expected. I've put in a cast to prevent warnings from picky compilers.
8249 34. If a DNS black list lookup found a CNAME record, but there were no A
8250 records associated with the domain it pointed at, Exim crashed.
8252 35. If a DNS black list lookup returned more than one A record, Exim ignored
8253 all but the first. It now scans all returned addresses if a particular IP
8254 value is being sought. In this situation, the contents of the
8255 $dnslist_value variable are a list of all the addresses, separated by a
8258 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
8259 to accept a host name and all its aliases if the forward lookup for any of
8260 them yielded the IP address of the incoming connection. Now it accepts only
8261 those names whose forward lookup yields the correct IP address. Any other
8262 names are discarded. This closes a loophole whereby a rogue DNS
8263 administrator could create reverse DNS records to break through a
8264 wildcarded host restriction in an ACL.
8266 37. If a user filter or a system filter that ran in a subprocess used any of
8267 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
8268 the wrong values were passed to the pipe command ($thisaddress had the
8269 value of $0, $0 had the value of $1, etc). This bug was introduced by
8270 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
8272 38. Improved the line breaking for long SMTP error messages from ACLs.
8273 Previously, if there was no break point between 40 and 75 characters, Exim
8274 left the rest of the message alone. Two changes have been made: (a) I've
8275 reduced the minimum length to 35 characters; (b) if it can't find a break
8276 point between 35 and 75 characters, it looks ahead and uses the first one
8277 that it finds. This may give the occasional overlong line, but at least the
8278 remaining text gets split now.
8280 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
8281 file descriptors might be low, and that setting 1000 would always raise it.
8282 It turns out that in some environments, the limit is already over 1000 and
8283 that lowering it causes trouble. So now Exim takes care not to decrease it.
8285 40. When delivering a message, the value of $return_path is set to $sender_
8286 address at the start of routing (routers may change the value). By an
8287 oversight, this default was not being set up when an address was tested by
8288 -bt or -bv, which affected the outcome if any router or filter referred to
8291 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
8292 log only when "message" or "log_message" are set. However, if one of the
8293 conditions was an address verification, or a call to a nested ACL, the
8294 messages generated by the underlying test were being passed through. This
8295 no longer happens. The underlying message is available in $acl_verify_
8296 message for both "message" and "log_message" expansions, so it can be
8297 passed through if needed.
8299 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
8300 new expansion $bh_ to give the encoded byte string without charset
8301 translation. Translation happens only if iconv() is available; HAVE_ICONV
8302 indicates this at build time. HEADERS_CHARSET gives the charset to
8303 translate to; headers_charset can change it in the configuration, and
8304 "headers charset" can change it in an individual filter file.
8306 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
8307 that creates RFC 2047 encoded "words" labels them as that charset instead
8308 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
8309 expansion operator; (ii) when Exim creates a From: line for a local
8310 message; (iii) when a header line is rewritten to include a "phrase" part.
8312 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
8313 buggy, causing it to skip the first lines of messages whose message ID
8314 ended in 'D'. This would not have bitten before Exim release 4.14, because
8315 message IDs were unlikely to end in 'D' before then. The effect was to have
8316 incorrect size information for certain domains.
8318 45. #include "config.h" was missing at the start of the crypt16.c module. This
8319 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
8322 46. If there was a timeout during a "random" callout check, Exim treated it as
8323 a failure of the random address, and carried on sending RSET and the real
8324 address. If the delay was just some slowness somewhere, the response to the
8325 original RCPT would be taken as a response to RSET and so on, causing
8326 mayhem of various kinds.
8328 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
8329 when I implemented it. It didn't allow for the fact that some option values
8330 may legitimately be negative (e.g. size_addition), and it didn't even do
8331 the right test for positive values.
8333 48. Domain names in DNS records are case-independent. Exim always looks them up
8334 in lower case. Some resolvers return domain names in exactly the case they
8335 appear in the zone file, that is, they may contain uppercase letters. Not
8336 all resolvers do this - some return always lower case. Exim was treating a
8337 change of case by a resolver as a change of domain, similar to a widening
8338 of a domain abbreviation. This triggered its re-routing code and so it was
8339 trying to route what was effectively the same domain again. This normally
8340 caused routing to fail (because the router wouldn't handle the domain
8341 twice). Now Exim checks for this case specially, and just changes the
8342 casing of the domain that it ultimately uses when it transmits the message
8345 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
8348 50. If a filter generated a file delivery with a non-absolute name (possible if
8349 no home directory exists for the router), the forbid_file option was not
8352 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
8353 the existing equality matching.
8355 52. Exim was using ints instead of ino_t variables in some places where it was
8356 dealing with inode numbers.
8358 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
8359 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
8360 called TMPDIR, and if it finds it is different, it changes its value.
8362 54. The smtp_printf() function is now made available to local_scan() so
8363 additional output lines can be written before returning. There is also an
8364 smtp_fflush() function to enable the detection of a dropped connection.
8365 The variables smtp_input and smtp_batched_input are exported to
8368 55. Changed the default runtime configuration: the message "Unknown user"
8369 has been removed from the ACL, and instead placed on the localuser router,
8370 using the cannot_route_message feature. This means that any verification
8371 failures that generate their own messages won't get overridden. Similarly,
8372 the "Unrouteable address" message that was in the ACL for unverifiable
8373 relay addresses has also been removed.
8375 56. Added hosts_avoid_esmtp to the smtp transport.
8377 57. The exicyclog script was not checking for the esoteric option
8378 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
8379 will work only if exicyclog is run under the appropriate euid.
8381 58. Following a discussion on the list, the rules by which Exim recognises line
8382 endings on incoming messages have been changed. The -dropcr and drop_cr
8383 options are now no-ops, retained only for backwards compatibility. The
8384 following line terminators are recognized: LF CRLF CR. However, special
8385 processing applies to CR:
8387 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
8388 nor a local message in the state where . is a terminator.
8390 (ii) If a bare CR is encountered in a header line, an extra space is added
8391 after the line terminator so as not to end the header. The reasoning
8392 behind this is that bare CRs in header lines are most likely either
8393 to be mistakes, or people trying to play silly games.
8395 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
8396 was being incorrectly given as 18 bytes larger than it should have been.
8397 This is a VOB (very old bug).
8399 60. This may never have affected anything current, but just in case it has:
8400 When the local host is found other than at the start of a list of hosts,
8401 the local host, those with the same MX, and any that follow, are discarded.
8402 When the list in question was part of a longer list of hosts, the following
8403 hosts (not currently being processed) were also being discarded. This no
8404 longer happens. I'm not sure if this situation could ever has previously
8407 61. Added the "/MX" feature to lists of hosts in the manualroute and query
8410 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
8411 header. This is something that is recommended in a new Internet Draft, and
8412 is something that is documented as being done by Sendmail. There are two
8413 possible values. For messages generated by the autoreply transport, Exim
8416 Auto-Submitted: auto-replied
8418 whereas for all other generated messages (e.g. bounces) it adds
8420 Auto-Submitted: auto-generated
8422 63. The "personal" condition in filters now includes a test for the
8423 Auto-Submitted: header. If it contains the string "auto-" the message it
8424 not considered personal.
8426 64. Added rcpt_include_affixes as a generic transport option.
8428 65. Added queue_only_override (default true).
8430 66. Added the syslog_duplication option.
8432 67. If what should have been the first header line of a message consisted of
8433 a space followed by a colon, Exim was mis-interpreting it as a header line.
8434 It isn't of course - it is syntactically invalid and should therefore be
8435 treated as the start of the message body. The misbehaviour could have
8436 caused a number of strange effects, including loss of data in subsequent
8437 header lines, and spool format errors.
8439 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
8440 client host had authenticated. This control can now be exercised by an ACL
8441 for more flexibility.
8443 69. By default, callouts do not happen when testing with -bh. There is now a
8444 variant, -bhc, which does actually run the callout code, including
8445 consulting and updating the callout cache.
8447 70. Added support for saslauthd authentication, courtesy of Alexander
8450 71. If statvfs() failed on the spool or log directories while checking their
8451 size for availability, Exim confusingly gave the error "space shortage".
8452 Furthermore, in debugging mode it crashed with a floating point exception.
8453 These checks are done if check_{spool,log}_{space,inodes} are set, and when
8454 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
8455 serious problem, Exim now writes to the main and panic logs when this
8456 happens, with details of the failure. It then refuses to accept the
8457 incoming message, giving the message "spool directory problem" or "log
8458 directory problem" with a 421 code for SMTP messages.
8460 72. When Exim is about to re-exec itself, it ensures that the file descriptors
8461 0, 1, and 2 exist, because some OS complain for execs without them (see
8462 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
8463 descriptors. However, the code omitted to check that the open succeeded,
8464 causing mysterious errors if for some reason the permissions on /dev/null
8465 got screwed. Now Exim writes a message to the main and panic logs, and
8466 bombs out if it can't open /dev/null.
8468 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
8469 interact so that it is all more flexible. It is supposed to remain
8470 backwards compatible. Also added extra_local_interfaces.
8472 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
8473 to bomb out with an assertion failure - to the client this appears as a
8474 connection drop. This problem occurs in the part of the code that was taken
8475 from the Samba project. Fortunately, the assertion is in a very simple
8476 function, so I have fixed this by reproducing the function inline in the
8477 one place where it is called, and arranging for authentication to fail
8478 instead of killing the process with assert().
8480 75. The SPA client code was not working when the server requested OEM rather
8481 than Unicode encoding.
8483 76. Added code to make require_files with a specific uid setting more usable in
8484 the case where statting the file as root fails - usually a non-root-mounted
8485 NFS file system. When this happens and the failure is EACCES, Exim now
8486 forks a subprocess and does the per-uid checking as the relevant uid.
8488 77. Added process_log_path.
8490 78. If log_file_path was not explicitly set, a setting of check_log_space or
8491 check_log_inodes was ignored.
8493 79. If a space check for the spool or log partitions fails, the incident is now
8494 logged. Of course, in the latter case the data may get lost...
8496 80. Added the %p formatting code to string_format() so that it can be used to
8497 print addresses in debug_print(). Adjusted all the address printing in the
8498 debugging in store.c to use %p rather than %d.
8500 81. There was a concern that a line of code in smtp_in.c could overflow a
8501 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
8502 initially expressed, the concern was not well-founded, because trailing
8503 spaces are removed early. However, if the trailing spaces were followed by
8504 a NULL, they did not get removed, so the overflow was possible. Two fixes
8507 (a) I re-wrote the offending code in a cleaner fashion.
8508 (b) If an incoming SMTP command contains a NULL character, it is rejected
8511 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
8512 runs initgroups(), so that if the Exim user is in any additional groups,
8513 they will be used during message reception.
8519 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob